Persistent air leak is a common complication of spontaneous pneumothorax being treated by chest tube drainage. In one report, the median time for spontaneous cessation of air leak was 7 days for primary spontaneous pneumothorax and 11 days for secondary spontaneous pneumothorax.1 In another report, 75% and 100% of cases of primary spontaneous pneumothorax resolved by 7 days and 14 days, respectively, whereas the corresponding proportions for secondary spontaneous pneumothorax were 61% and 79%, respectively.2 It is generally recommended that surgical intervention be considered when an air leak continues for 7 to 14 days after chest tube insertion. Unfortunately, some patients are poor candidates for surgery, and some patients may refuse surgery. Prolonged chest tube in situ, however, is undesirable because of an increased risk of complications, prolonged immobilisation and its consequences, and increased duration of hospital stay. Alternative means to shorten the duration of air leakage are thus needed.

Endobronchial one-way valves (EBVs) have been used to treat persistent air leak complicating pneumothorax. The rationale is that air leakage will stop if air is prevented from entering the airway leading to the leak site. Preliminary case reports showed encouraging results.3 4 5 6 7 8 9 10 11 12 13 14 Larger case series showed similarly favourable results.15 16 17 18 19 20 21 22 23 However, most of these reports include a mix of post-surgical pneumothoraces, spontaneous pneumothoraces, and other aetiologies. Moreover, there is scanty information on overall efficacy, short-term safety, and factors related to favourable clinical outcomes.

The Princess Margaret Hospital, an acute-care general hospital in Hong Kong, has been using EBV for persistent air leak complicating pneumothorax on compassionate grounds since 2008, and its preliminary experience on three apparently successful cases was reported in 2009.11 However, it was subsequently found that in many cases, the leak site could not be identified, so EBV could not be deployed. In some other cases, despite leak site identification and EBV deployment, the air leak may not resolve completely, or it may recur after a few hours. We therefore hypothesise that EBV treatment has a low success rate in real-life situations. To test this hypothesis, we retrospectively compared the clinical course of patients with and without EBV use for persistent air leak complicating spontaneous pneumothorax during an 8-year period at a single centre.

This retrospective chart review was based on patients who were managed at the Princess Margaret Hospital from May 2008 to April 2016. Eligible patients were those with spontaneous pneumothorax who were undergoing chest tube drainage and who had had an air leak lasting more than 1 week. All showed bubbling in the drainage bottle during both expiration and inspiration, with water suction at -20 cm. Patients either refused surgery or were considered by the thoracic surgeon to be unsuitable for surgical intervention owing to advanced age, poor lung function, major medical co-morbidities, or a combination of these.

Bronchoscopy with the intention of EBV implantation to stop the air leak was suggested to eligible patients. Some were reluctant at first but consented after further days of air leakage. Procedures for identifying the leak site and deploying EBV were as described previously.11 Briefly, an endobronchial inflatable balloon was used to occlude lobar or segmental bronchi sequentially. If bubbling from the chest drainage bottle stopped, or was greatly reduced, then the lobe or segment was selected for endobronchial valve implantation. The Zephyr EBV (Emphasys Inc, now Pulmonx Inc, Redwood City [CA], United States) was used throughout. Patients were observed closely for continued air leak after bronchoscopy. Chest tubes were removed as soon as possible after air-leak cessation, typically within 1 or 2 days, as is usual clinical practice.

Data on demographic and disease characteristics, details of bronchoscopy, and survival after bronchoscopy were recorded. Spirometry data were analysed if available within 1 year of bronchoscopy. Spirometry was performed according to the American Thoracic Society / European Respiratory Society criteria24 and using reference values for Hong Kong adults.25 No spirometry was performed during the period of chest tube drainage or with EBV in situ. Computed tomography images, if available, were viewed in axial, sagittal, and coronal planes by a single radiologist who was blinded to other study data. The integrity of the fissures was defined as more than 90% completeness on at least one axis.26 Outcome measures included whether EBV was implanted, time to air-leak cessation, and 30-day all-cause mortality after bronchoscopy.

Therapeutic success was defined as cessation of the air leak within 72 hours of EBV implantation. Cases with EBV implanted but with the air leak lasting more than 72 hours afterwards, cessation of air leak within 72 hours of bronchoscopy owing to other interventions (such as surgery), cases without EBV implantation, and cases of chest tube removal while the air leak persisted (such as following unintended displacement) were considered cases of failure.

The Mann-Whitney U test and Fisher’s exact test were used to examine differences between groups. The Kaplan-Meier log-rank test and Gehan-Breslow-Wilcoxon test were used to analyse time to air-leak cessation after first bronchoscopy among patients with and without EBV implantation. The association between EBV implantation and air-leak cessation was assessed in a multivariable Cox proportional hazards regression model that was adjusted for ‘days on chest tube before first bronchoscopy’ and factors with P<0.2 in univariable regression analyses. The accuracy, sensitivity, specificity, and positive and negative predictive values of using the presence of an intact interlobar fissure as an indicator of a successful outcome were calculated. Statistical analyses were performed using SPSS 22.0 for Windows (IBM Corp., Armonk [NY], United States) and OpenEpi: Open Source Epidemiologic Statistics for Public Health, version 3.01 (http://www.openepi.com). Statistical significance was set at P<0.05.

This study was approved by the Research Ethics Committee of the Kowloon West Cluster of the Hong Kong Hospital Authority, with the requirement for patient consent waived. This report conforms to the STROBE 2008 guidelines.27

During the study period, 38 patients underwent bronchoscopy for persistent air leak complicating spontaneous pneumothorax at our institution. One patient was excluded because the air leak was subsequently found to result from the chest tube having been inserted into an airway. Of the remaining 37 patients, 35 were assessed by the surgeon and considered unfit for surgery, and two patients with primary spontaneous pneumothorax refused surgery. The median patient age was 71 years and most (34; 92%) were males. The majority (83%) were ever-smokers. All but one (97%) had at least one co-morbidity and the median Charlson co-morbidity index was 1. Fifteen patients (41%) had a left pneumothorax. Thirty-three (89%) had secondary spontaneous pneumothorax, of which 23 had chronic obstructive pulmonary disease with or without other lung diseases. Of the 22 secondary spontaneous pneumothorax cases with spirometric data, the median forced expiratory volume in 1 second (FEV₁) was 0.94 L; median percentage of predicted FEV₁ was 43.5%; and median FEV₁ to forced vital capacity ratio was 0.45. Eleven patients (30%) had at least one attempt at talc pleurodesis before bronchoscopy. The median number of days off chest tube use before bronchoscopy was 25 (Table 1).

Nineteen patients (51%) had a mean of 2.6 EBVs implanted (range, 1-6). The sites of EBV implantation were as follows: 11 in the right upper lobe, one in the right middle lobe, none in the right lower lobe, five in the left upper lobe, and two in the left lower lobe. Of the 18 patients without an EBV implanted, the target site for EBV implantation could not be identified in 17. In the remaining patient, despite identification of the air leak in the left upper lobe, EBV implantation was followed by severe oxygen desaturation and had to be abandoned.

There were eight successful cases among patients with EBVs implanted. Among these, seven patients had immediate and lasting cessation of the air leak, and the chest tube was removed within 2 days. One patient had immediate air-leak cessation but there was recurrence after 2 hours. The air leak subsided completely within 72 hours, and the chest tube was removed on day 4. Among the 11 cases of failure after EBV implantation, the air leak stopped immediately in three cases but recurred soon after and persisted beyond 72 hours. In the other eight cases, the air leak was reduced temporarily but was present beyond 72 hours. In the group without an implanted EBV, the earliest time for air-leak cessation was day 5, and the chest tube was removed on day 8. In a Kaplan-Meier comparison of the EBV and no-EBV groups in the number of days from first bronchoscopy to air-leak cessation, the EBV group did better in the first 30 days, but the no-EBV group caught up by day 45. There was a statistically significant difference between the two groups according to the Gehan-Breslow-Wilcoxon test (P=0.027), but not the log-rank test (P=0.138) [Fig Part a]. When the same comparison was done with the eight successful cases removed from the EBV group, the two curves overlapped throughout and there was no statistically significant difference between the two groups (P=0.881 by Gehan-Breslow-Wilcoxon test and P=0.976 by log-rank test) [Fig Part b].

In a comparison of demographic and clinical characteristics between the EBV group and the no-EBV group, the former had a significantly poorer FEV₁, but not percentage of predicted FEV₁. This group also had the chest tube in place for a significantly longer duration before bronchoscopy (Table 1).

We further examined factors related to air-leak cessation after first bronchoscopy by Cox proportional hazards regression analysis. The implantation of an EBV was significantly associated with air-leak cessation (adjusted hazard ratio=2.39, 95% CI=1.13-5.05; P=0.023), whereas the Charlson co-morbidity index was significantly associated with delayed air-leak cessation (adjusted hazard ratio=0.78, 95% CI=0.63-0.97; P=0.026). The number of days of chest tube use before first bronchoscopy was not associated with air-leak cessation (Table 2).

When comparing the eight successful cases with the 11 failed cases among those with implanted EBVs, we found no difference in any of the demographic or disease variables between the two groups, apart from the Charlson co-morbidity index, which was significantly lower for the success group (Table 3). Additionally, all eight patients (100%) in the success group had air-leak cessation immediately after EBV implantation, versus only three of 11 (27%) in the failure group (P=0.003). When we compared the eight successful cases with all 29 failed cases, again the median (interquartile range) Charlson co-morbidity index was significantly lower for the success group: 1 (0.25-1) vs 2 (1-3) [P=0.029]. This group also had significantly poorer FEV₁, but not in percentage of predicted FEV₁, and showed a trend towards more severe airflow obstruction, although this did not reach statistical significance (Table 3).

Twenty-three patients had thoracic computed tomography performed to examine the integrity of interlobar fissures. For the 13 patients with all fissures intact, six had an EBV implanted and were classed as successful cases, four had EBV implanted but were classed as failed cases, and three did not have an EBV implanted. There were five patients each in the group with any non-intact ipsilateral fissure and in the group with any non-intact contralateral fissure. Both groups had two patients with an implanted EBV and three without, and all experienced treatment failure.

When using the presence of all intact fissure(s) to screen for successful EBV treatment among the 23 patients who had undergone thoracic computed tomography, the accuracy was 69.6%, the sensitivity and negative predictive value were both 100%, and the specificity and positive predictive value were 58.8% and 46.2%, respectively (Table 4). When the criterion for success was immediate air-leak cessation after EBV implantation, the accuracy among the 19 patients with an implanted EBV was 84.2%, the sensitivity and negative predictive value remained at 100%, and both the specificity and positive predictive value were 72.7% (Table 4).

There was no incident of valve displacement, bleeding, or post-obstructive pneumonia. Three patients died within 30 days of EBV implantation and all had advanced age and multiple co-morbidities (Table 5). In two patients, the causes of death were clearly related to ongoing pre-existing disease. One patient had a sudden cardiac arrest on day 29 after being successfully treated with EBV, although the relationship of death to EBV was uncertain. The earliest death in the no-EBV group occurred on day 43 after bronchoscopy and the cause of death was lung cancer.

Three patients underwent a second bronchoscopy after the first one failed, with one success and two failures. Sixteen patients received talc pleurodesis (median, 2 times; range, 1-5 times). One young patient with primary spontaneous pneumothorax underwent surgery 2 days after failed identification of the leak site by bronchoscopy. Seventeen patients did not receive any further treatment, mostly owing to early air-leak cessation.

Thirteen patients had their implanted EBVs removed after a median of 43 days (range, 21-155 days). For the remaining six patients, three died within 30 days before EBV removal was considered. Another patient had a second bronchoscopy with an additional EBV implanted but died within 30 days of the second bronchoscopy. The fifth patient had advanced lung cancer and removal of EBV was deemed unnecessary; she died 9 months after bronchoscopy. The sixth patient had severe chronic obstructive pulmonary disease and there was subjective improvement of respiratory and health status after EBV implantation, so it was decided that the EBV should remain in situ indefinitely.

We have reported the first real-life cohort study that consisted entirely of spontaneous pneumothorax cases with persistent air leak treated with EBV. Our results confirm those of previous reports that EBV can be useful in hastening air-leak cessation in patients with this condition. Nonetheless, it also highlights the fact that failures are common. A site for EBV implantation was not identified in nearly half of the patients. For the 19 patients with an EBV implanted, only eight (42%) had a clear success. The overall success rate was thus only 22%.

In this retrospective study, we attempted to assess the efficacy of EBV treatment by comparing those who had an EBV implanted with those in whom EBV implantation was denied owing to inability to identify an implantation site or lung function that was too poor. This design is obviously inferior to a prospective randomised design. Nevertheless, the two study groups had a similar number of patients who were similar in many demographic and disease aspects. The Kaplan-Meier comparison suggested that EBV treatment is efficacious in hastening air-leak cessation. The significantly longer duration of air leakage before first bronchoscopy for the EBV group is a potential confounder, because the longer the air leak exists, the higher the probability that spontaneous resolution will occur. Nonetheless, the Cox hazards regression analysis suggested that it was not a significant factor. This finding also lends further support to the efficacy of EBV implantation in enabling early air-leak cessation.

A definition of success for using EBV implantation to treat persistent air leak complicating pneumothorax is difficult to formulate and has seldom been discussed in previous studies. Our criteria of cessation of air leak within 72 hours was arbitrary but is supported by our observation that patients who did not meet the criteria behaved almost exactly as if no EBV was implanted. This finding suggests that for cases in which EBV is implanted but fails, other forms of treatment should be sought early.

One interesting finding is that higher co-morbidity burden seemed to be a risk factor for delayed air-leak cessation, irrespective of EBV implantation status. Delayed cessation of air leak may be translated as delayed healing of the lesion responsible for the air leak, which is expected in patients with more co-morbidities and who are thus usually sicker. Nonetheless, our study was not designed to examine this association, and a properly designed study would be needed to provide definitive answers.

It has been widely accepted that fissure completeness of the target lobe is strongly correlated with significant lobar collapse after implantation of EBVs for volume reduction in severe pulmonary emphysema.28 29 The same is probably true for EBV treatment of persistent air leak complicating pneumothorax, although as far as we are aware there are no published data on this. Not surprisingly, our results showed that intact interlobar fissures were a necessary, but not sufficient, condition for a successful outcome. However, an interesting and unexpected finding is that patients with a non-intact fissure in the contralateral lung, but intact fissure in the ipsilateral lung, behaved similarly to those with a non-intact fissure in the ipsilateral lung. We have no explanation for this, and these findings need to be confirmed with larger studies.

Immediate cessation of air leak after completion of EBV implantation was strongly predictive of a successful outcome. Importantly, all eight patients with an implanted EBV and without immediate cessation of air leak failed to respond to treatment. In such cases, the EBV should have been removed immediately to save costs and to avert possible adverse events associated with EBVs. A further implication is that stringent balloon testing of cessation of air leak should be performed and, in the presence of any uncertainty, EBVs should not be implanted.

In all our patients, we encountered no adverse events directly attributable to EBVs. Nonetheless, the three early deaths within 30 days of EBV implantation is worrying. The causes of death in two cases were clearly the severe pre-existing illness, but the possibility that recent EBV implantation hastened the terminal event cannot be excluded. The case of sudden cardiac death raises suspicion that the implanted EBV was implicated. More data are needed to determine whether EBV implantation is associated with increased early mortality.

There were several limitations to our study. First, it was a retrospective one, so some data were unavailable. Second, it was not a randomised controlled study. Third, it was a single-centre study, making the data less generalisable to a wider setting. Still, single-centre retrospective case series do have the advantage that practices are more uniform and results easier to interpret. Fourth, our case series was small, making it difficult to identify definitive factors associated with clinical outcomes. Fifth, some groups had a very small number of patients, and especially of females, patients with primary spontaneous pneumothorax, and patients with a lesion in the lower lobes. Finally, chart review and data collection (besides radiological data) were performed by an investigator who was not blinded to the study outcomes, and this may have been a source of bias.

We conclude that EBV implantation via the flexible bronchoscope can be useful in hastening air-leak cessation in patients with persistent air leak complicating spontaneous pneumothorax. Only about one-fifth of subjects, however, showed unequivocal benefit, and safety of this form of treatment needs further evaluation. Bilateral intact interlobar fissures seem to be a necessary, though not sufficient, condition for treatment success, and patients with fewer medical co-morbidities and immediate air-leak cessation after completion of EBV implantation seem to have a higher likelihood of treatment success. Further evaluation by randomised controlled trials is warranted.

We thank the doctors and nurses of the Respiratory Team, Princess Margaret Hospital, for taking care of the patients, and nurses of the Central Endoscopy Unit, Princess Margaret Hospital, for providing technical support for the bronchoscopy procedures.

The authors have no conflicts of interest to disclose.

1. Mathur R, Cullen J, Kinnear WJ, Johnston ID. Time course of resolution of persistent air leak in spontaneous pneumothorax. Respir Med 1995;89:129-32. Crossref

2. Chee CB, Abisheganaden J, Yeo JK, et al. Persistent air-leak in spontaneous pneumothorax—clinical course and outcome. Respir Med 1998;92:757-61. Crossref

3. Snell G, Holsworth L, Fowler S, et al. Occlusion of a broncho-cutaneous fistula with endobronchial one-way valves. Ann Thorac Surg 2005;80:1930-2. Crossref

4. Ferguson JS, Sprenger K, Van Natta T. Closure of a bronchopleural fistula using bronchoscopic placement of an endobronchial valve designed for the treatment of emphysema. Chest 2006;129:479-81. Crossref

5. Mitchell KM, Boley TM, Hazelrigg SR. Endobronchial valves for treatment of bronchopleural fistula. Ann Thorac Surg 2006;81:1129-31. Crossref

6. Feller-Kopman D, Bechara R, Garland R, Ernst A, Ashiku S. Use of a removable endobronchial valve for the treatment of bronchopleural fistula. Chest 2006;130:273-5. Crossref

7. Fann JI, Berry GJ, Burdon TA. The use of endobronchial valve device to eliminate air leak. Respir Med 2006;100:1402-6. Crossref

8. Anile M, Venuta F, De Giacomo T, et al. Treatment of persistent air leakage with endobronchial one-way valves. J Thorac Cardiovasc Surg 2006;132:711-2. Crossref

9. Toma TP, Kon OM, Oldfield W, et al. Reduction of persistent air leak with endoscopic valve implants. Thorax 2007;62:830-3. Crossref

10. Santini M, Fiorelli A, Vicidomini G, Laperuta P, Di Crescenzo VG. Iatrogenic air leak successfully treated by bronchoscopic placement of unidirectional endobronchial valves. Ann Thorac Surg 2010;89:2007-10. Crossref

11. Yu WC, Yeung YC, Chang Y, et al. Use of endobronchial one-way valves reveals questions on etiology of spontaneous pneumothorax: report of three cases. J Cardiothorasc Surg 2009;4:63. Crossref

12. Schiavon M, Marulli G, Zuin A, et al. Endobronchial valve for secondary pneumothorax in a severe emphysema patient. Thorac Cardiovasc Surg 2011;59:509-10. Crossref

13. El-Sameed Y, Waness A, Al Shamsi I, Mehta AC. Endobronchial valves in the management of broncho-pleural and alveolo-pleural fistulae. Lung 2012;190:347-51. Crossref

14. Ambrosino N, Ribechini A, Allidi F, Gabbrielli L. Use of endobronchial valves in persistent air leaks: a case report and review of the literature. Expert Rev Respir Med 2013;7:85-90. Crossref

15. Travaline JM, McKenna RJ Jr, De Giacomo T, et al. Treatment of persistent pulmonary air leaks using endobronchial valves. Chest 2009;136:355-60. Crossref

16. Gillespie CT, Sterman DH, Cerfolio RJ, et al. Endobronchial valve treatment for prolonged air leaks of the lung: a case series. Ann Thorac Surg 2011;91:270-3. Crossref

17. Mahajan AK, Verhoef P, Patel SB, Carr G, Kyle Hogarth D. Intrabronchial valves: a case series describing a minimally invasive approach to bronchopleural fistulas in medical intensive care unit patients. J Bronchol Intervent Pulmonol 2012;19:137-41. Crossref

18. Firlinger I, Stubenberger E, Müller MR, Burghuber OC, Valipour A. Endoscopic one-way valve implantation in patients with prolonged air leak and the use of digital air leak monitoring. Ann Thorac Surg 2013;95:1243-9. Crossref

19. Dooms CA, Decaluwe H, Yserbyt J, De Leyn P, Van Raemdonck D, Ninane V. Bronchial valve treatment for pulmonary air leak after anatomical lung resection for cancer. Eur Respir J 2014;43:1142-8. Crossref

20. Cordovilla R, Torracchi AM, Novoa N, et al. Endobronchial valves in the treatment of persistent air leak, an alternative to surgery [in English, Spanish]. Arch Bronconeumol 2015;51:10-5. Crossref

21. Reed MF, Gilbert CR, Taylor MD, Toth JW. Endobronchial valves for challenging air leaks. Ann Thorac Surg 2015;100:1181-6. Crossref

22. Hance JM, Martin JT, Mullett TW. Endobronchial valves in the treatment of persistent airleaks. Ann Thorac Surg 2015;100:1780-6. Crossref

23. Gilbert CR, Casal RF, Lee HJ, et al. Use of one-way intrabronchial valves in air leak management after tube thoracostomy drainage. Ann Thorac Surg 2016;101:1891-6. Crossref

24. Miller MR, Hankinson J, Bursasco V, et al. Standardisation of spirometry. Eur Respir J 2005;26:319-38. Crossref

25. Ip MS, Ko FW, Lau AC, et al. Updated spirometric reference values for adult Chinese in Hong Kong and implications on clinical utilization. Chest 2006;129:384-92. Crossref

26. Sciurba FC, Ernst A, Herth FJ, et al. A randomized study of endobronchial valves for advanced emphysema. N Engl J Med 2010;363:1233-44. Crossref

27. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008;61:344-9. Crossref

28. Koster TD, van Rikxoort EM, Huebner RH, et al. Predicting lung volume reduction after endobronchial valve therapy is maximized using a combination of diagnostic tools. Respiration 2016;92:150-7. Crossref

29. Schuhmann M, Raffy P, Yin Y, et al. Computed tomography predictors of response to endobronchial valve lung reduction treatment. Comparison with Chartis. Am J Respir Crit Care Med 2015;191:767-74. Crossref

Audits of surgical mortality are used worldwide to monitor surgical outcome and achieve quality assurance.1 By measuring and comparing properly collected, risk-adjusted surgical outcome data, quality of surgical care could be enhanced in participating institutions.2 It has been demonstrated in several countries that adoption of a national surgical audit programme can reduce mortality.2 3

The Hong Kong Hospital Authority (HA) was established in 1991. It is a government statutory body responsible for the management of 42 public hospitals and institutions, 47 specialist out-patient clinics, and 73 general out-patient clinics in Hong Kong. Seventeen HA hospitals have surgical departments; all of them provide an elective surgery service and 14 also provide an emergency surgical service. In 2016, over 80% of all hospital admissions in Hong Kong were under the management of the HA.4 Therefore, it is important for the HA to develop tools with which to measure and improve performance. For this purpose, the HA Coordinating Committee of Surgery set up a Central Surgical Audit Unit in 2001. From 2002 to 2007, the unit conducted clinical audits based on retrospective cumulative data to compare the performance of the 17 surgical departments. One to two major or ultra-major operations, such as major hepatectomy, oesophagectomy, and major lung resection, were selected each year for comparison. The risk-adjustment model used was based on the Physiological and Operative Severity Score for Enumeration of Mortality and Morbidity (POSSUM).5 6 Those audits were procedure-based and mainly focused on mortality, and hence a more comprehensive risk-adjusted outcome audit was needed.

In 2008, a new programme—namely, the Surgical Outcomes Monitoring and Improvement Programme (SOMIP)—was launched by the HA with the aim of monitoring and improving surgical quality. The programme was governed by a steering committee comprising surgeons, an anaesthetist, a physician, administrative managers, and statisticians. Risk-adjusted mortality and morbidity rates of elective and emergency major and ultra-major operations at each hospital are measured and reported yearly. An annual forum is held to disseminate the results and allow the sharing of best practices.

This study aimed to assess the changes in overall surgical outcomes for patients after implementation of SOMIP.

Between July 2008 and June 2016, SOMIP captured data of all elective and emergency major/ultra-major operations (except those in children younger than 1 year) that were performed by general surgery, urology, plastic and reconstructive surgery, and paediatric surgery teams at all 17 HA hospitals. A total of 230 variables were collected from each patient: 10 patient demographic variables, 83 preoperative and operative variables, 31 laboratory test results, and 40 postoperative events and 66 postoperative adverse events in the first 30 days after surgery. Demographic data and laboratory test results were mostly automatically retrieved from various HA clinical information systems. For data that required manual retrieval, six full-time SOMIP nurse reviewers were employed by the HA head office for this purpose. Preoperative and operative variables, as well as postoperative complications occurring up to 30 days after the index operation, were retrieved from patient records by the SOMIP nurse reviewers. Mortality at 30 days, 60 days, and 90 days were also retrieved from the HA electronic database. These data were endorsed and submitted by each surgical department’s surgical supervisor within 60 days of surgery.

Both the manually captured and automatically captured data were entered into a tailor-made SOMIP electronic database. Data variable definitions were listed in the operation manual of the programme. To ensure data validity and consistency, all nurse reviewers completed comprehensive training on data definition and criteria, and regular nurse reviewer meetings were held to clarify any queries. All data were endorsed by the surgical supervisor of the respective surgical department. When necessary, data definitions were modified.

An inter-rater reliability test was performed each May and completed within a month so as to ensure consistency among nurse reviewers as well as data accuracy. Fifty cases were sampled for evaluation using a stratified systematic sampling method. For those selected cases, an independent nurse from the SOMIP team repeated the data collection using a designated data template, without prior knowledge of the information recorded by the original nurse reviewer. Data quality was measured by comparing two sets of data, and inferred by a score defined as the percentage of agreement between nurse reviewer and the SOMIP working team for each data item. The mean score of all data items was used to assess overall performance. The overall result was satisfactory and the mean score of all data items was 99.3% (range of individual item scores, 95.2%-100%).

Among the variables collected, preoperative risk factors including demographic data; general health and lifestyle variables; and major respiratory, cardiovascular, hepatobiliary, renal, vascular, central nervous system, and immune co-morbidities were deemed particularly important. These risk factors were modified from those in the American College of Surgeons National Surgical Quality and Improvement Program (NSQIP) to suit the local context.7 8 9 Operative variables included intra-operative blood loss, American Society of Anesthesiologists physical status classification, procedure complexity score, surgical subspecialty, wound classification, operative magnitude, and operative time. Before analysis and reporting, SOMIP data were cleaned and verified by four surgeons and the SOMIP surgical supervisors of each hospital. Questionable cases were reviewed accordingly. Cases of liver transplantation, renal transplantation, multiple trauma, and major bowel ischaemia (Table 1)10 11 12 13 14 15 16 17 were excluded from the risk-adjusted model analysis.

Comparative risk-adjusted models (different models for emergency operations and elective operations) were generated using hierarchical logistic regression. The 30-day risk-adjusted mortality models expressed hospital performance as expected odds ratios. A risk-adjusted observed-to-expected mortality ratio (O/E ratio) was then calculated for each hospital. The O/E ratio is a quotient between the observed number of deaths and the expected number of deaths; the latter was calculated by a logistic regression method based on significant independent risk factors. Together with the 90% confidence intervals, O/E ratios were depicted by caterpillar plots and boxplots. Hospitals with the lower limit of the 90% confidence interval of O/E ratios greater than 1 were defined as ‘high outliers’; hospitals with the upper limit of the 90% confidence interval of O/E ratios lower than 1 were defined as ‘low outliers’. The risk-adjusted outcome of a ‘high-outlier’ hospital was probably worse than the average outcome, and that of a ‘low-outlier’ hospital was probably better than the average outcome.

In addition to risk-adjusted postoperative mortality, various general medical and surgical complications, as well as specific complications (anastomotic leakage, surgical site infection, acute myocardial infarction, pneumonia), were recorded. The list of complications recorded and the method to derive the SOMIP risk-adjustment model have been described in detail in the annual SOMIP Report.10 11 12 13 14 15 16 17 Different levels of confidence were used for different outcome variables—90% confidence interval for mortality rates, 95% confidence interval for major complications, and 99% confidence interval for morbidity rates. Using SOMIP data together with other useful information extracted from the HA Executive Information System (eg, bed occupancy, nursing manpower, intensive care unit support, and surgeon workload), root-cause analyses were performed using multilevel logistic regression, as described in the annual SOMIP Report.10 11 12 13 14 15 16 17

The discriminative power of the risk-adjusted models was measured by the C-index, area under the receiver-operating characteristic curve (AUC). The closer the C-index is to 1, the better the discriminative power of the model is: a C-index of ≥0.8 indicates excellent discriminative power.18 The calibration accuracy of the models was assessed by the Hosmer-Lemeshow goodness-of-fit test (HL test). The calibration of the model was rejected if P<0.05. The Mann-Kendall non-parametric trend test was used to identify trends (positive or negative) in the annual data series for both crude mortality rates (2009/2010 to 2015/2016) and risk-adjusted O/E ratios (2011/2012 to 2015/2016). A very high positive value of the Mann-Kendall statistic (S) indicated an increasing trend; a very low negative value indicated a decreasing trend. The test statistic Z-score was used as a measure of trend significance.

Age distribution trends are summarised in Table 2.10 11 12 13 14 15 16 17 From 2008/2009 to 2015/2016, the proportion of people aged 61-70 years increased by 7 percentage points (from 19% to 26%), whereas the proportions of people aged 41-50 years and 71-80 years decreased by 2 and 5 percentage points, respectively.

Table 310 11 12 13 14 15 16 17 shows proportions of patients taking regular medication for diabetes mellitus and hypertension before surgery from 2009/2010 to 2015/2016, as well as those currently smoking (within 1 year) and drinking more than 2 units of alcohol per day in the previous 2 weeks. Over 70% of patients had at least one of these four conditions, whereas about 40% had a history of regular use of hypertension medication before surgery.

Annual numbers of elective operations by specialty are summarised in Table 4.10 11 12 13 14 15 16 17 For the 10 listed elective operations, the most frequently performed were in urology, consistently constituting 28% of the caseload from 2008/09 to 2015/16. The least frequently performed procedures were parotid surgery (1%) and paediatric surgery (1%).

From July 2008 to June 2016, eight SOMIP reports were published.10 11 12 13 14 15 16 17 They showed that the HA overall crude mortality rate approximately halved over this time. The crude 30-day mortality rate for emergency operations dropped gradually from 10.8% in the year 2009/2010 to 5.6% in 2015/2016 (Fig 1). Similarly, the crude 30-day mortality rate for elective operations more than halved: from 0.9% in 2009/2010 to 0.4% in 2015/2016 (Fig 2).

In the 5-year cumulative comparison analysis (2011/2012 to 2015/2016), both models had excellent discriminative power and good calibration accuracy. For emergency operations, the AUC was >0.9 and the HL test statistic was >0.1; for elective operations, the AUC was >0.89 and the HL test statistic was >0.2. The risk-adjusted observed-to-expected 30-day mortality ratio for both types of surgery showed a similar downward trend to the crude mortality rates. For emergency operations, the risk-adjusted O/E ratios were 1.126, 1.022, 1.113, 0.921, and 0.796 across the 5 years (Fig 3). These values show a statistically significant downward (negative) trend (S= –0.8; P<0.05). The results of the Mann-Kendall analyses are summarised in Table 4. For elective operations, risk-adjusted O/E ratios were 1.150, 1.229, 0.881, 0.862, and 0.859 across the 5 years (Fig 4). These values show a statistically significant downward (negative) trend (S= –0.8; P<0.05) [Table 5].

Before the turn of the century, most hospital records in HA hospitals were handwritten and retained by individual hospitals. There was no convenient means by which to share patient details among hospitals. Around 2001, the HA implemented a number of clinical management electronic systems, such as the electronic patient record, Operation Theatre Record System, and Clinical Data Analysis and Reporting System, at all HA hospitals. By virtue of this infrastructure, patient records and information about diagnoses and operations could be accessed at a central level. Based on this central clinical database, a Quality Assurance Subcommittee under the Coordinating Committee of Surgery commenced small-scale comparative clinical audits for ultra-major operations in 2002, focusing on one to two ultra-major operations per year. The audits provided basic information about hospital performance for the selected operation, such as number of procedures, age distribution of patients, and mortality rate.

Subsequently, the Coordinating Committee of Surgery developed a more robust system to monitor more major operations at the same time. It was decided to follow the framework of NSQIP, which was developed in 1994 by the Veterans Affairs Hospitals in the United States to monitor risk-adjusted surgical operation outcomes. Like HA hospitals, Veterans Affairs Hospitals are managed by a central governing body and equipped with a comprehensive electronic medical records system. Studies19 20 showed a significant improvement in both mortality and morbidity over time and thus, in 2004, NSQIP was extended to private hospitals and endorsed by the American College of Surgeons. Using NSQIP as the blueprint, the HA launched SOMIP in 2008. The SOMIP adopted similar risk-adjustment variables, use of nurse reviewers to collect data, a focus on hospital performance rather than individual surgeon performance, and similar methods of data analysis and determination of outliers. Moreover, SOMIP allowed individual HA hospitals to benchmark their performance against the overall HA average.

As all surgical patients have a different health status, their operation outcomes will likewise differ. Appropriate adjustment for different patient risks is essential when interpreting hospital mortality rates. To adjust for different risk factors, over 100 patient risk factors were captured for each enrolled patient. For NSQIP, one risk-adjustment mortality model was constructed for all operations. In contrast to NSQIP, separate models were devised for emergency and elective operations in SOMIP.

Hospital outliers can be identified by O/E ratios if the confidence interval of the O/E ratio is greater than 1, meaning that after balancing the different risks of hospital patients, their clinical outcomes are most likely different from the rest. From the results of this study, it was encouraging to find a significant trend of reduction in crude mortality rates and O/E ratios for both elective and emergency operations at HA hospitals over the past 5 years. According to the significant reduction in expected odds ratios over the years, this improvement is genuine and not due to patient selection.

There are several possible reasons behind the changes: public identification as a poor performer is a strong incentive for change in HA hospitals; sharing best practices on perioperative patient care is an important educational activity that takes place annually through the SOMIP Forum; the HA Head Office invests more resources into deficient hospitals; and changing attitudes towards managing surgical complications by other colleagues from the intensive care unit are also helpful. All of these may have contributed to the change.

The SOMIP has a number of limitations. The coverage of SOMIP is not as complete as that of NSQIP, since many surgical departments such as orthopaedics and neurosurgery are not included. Monitoring is done by retrospective annual case collection (from 1 July to 30 June) because it takes 13 months to complete case enrolment and an additional 3 months for data verification, model building, and statistical analysis. Because of the small number of events, the current programme is not able to determine the risk-adjusted outcomes of individual operations or surgeons. Furthermore, because this programme relies heavily on the HA central electronic database, it is not easy to extend it to hospitals without this information infrastructure. Although the trend of reduction in mortality was statistically significant, we were not able to demonstrate a causal relationship with SOMIP implementation.

Potential issues with data quality may have affected the outcomes. In the inter-rater reliability test, the nurse reviewers were not blinded and this may have caused information bias. Also, quality of data collection in the initial 2 years may have been unreliable. As a result, the 5-year cumulative comparison analysis for emergency and elective operations commenced from 2011/2012, rather than 2008/2009. Furthermore, data definitions are updated regularly in the operation manual and could have affected the time trend analysis. Nonetheless, the SOMIP team considered changes in data definitions to be minor and did not expect a significant impact on the risk models.

Mortality could be influenced by many factors; ensuring risk adjustments are adequate and appropriate would be a challenge. Disease factors, stage of disease, and treatment options may not be fully taken into account by the risk-adjusted models, and data readiness and availability are further constraints. Surgeon skill and experience was another aspect that could not be accommodated and was difficult to adjust for. In the HA, surgical operations are performed by a team; therefore, it would be difficult to separate individual surgeon experience and credentials from those of the whole team.

From 2008 to 2016, the HA’s overall crude mortality rates and risk-adjusted O/E ratios showed a significant trend of reduction for both emergency and elective operations. The SOMIP enhances understanding of surgical performance and helps identify areas for improvement. It allows individual HA hospitals to benchmark their performance against the overall HA average through risk-adjusted O/E ratios.

We thank the SOMIP Steering Committee of the Hospital Authority, the Coordinating Committee of Surgery, and the Biostatistics team of The Chinese University of Hong Kong for their contributions and helpful comments on this manuscript.

The authors have no conflicts of interest to disclose.

1. Lui CW, Boyle FM, Wysocki AP, et al. How participation in surgical mortality audit impacts surgical practice. BMC Surg 2017;17:42. Crossref

2. Khuri SF, Daley J, Henderson W, et al. Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs surgical risk study. J Am Coll Surg 1997;185:315-27. Crossref

3. Kiermeier A, Babidge WJ, McCulloch GA, Maddern GJ, Watters DA, Aitken RJ. National surgical mortality audit may be associated with reduced mortality after emergency admission. ANZ J Surg 2017;87:830-6. Crossref

4. Hospital Authority Statistical Report (2015-16). Available from: http://www.ha.org.hk/haho/ho/stat/HASR15_16.pdf. Accessed 10 Nov 2017.

5. Copeland G, Jones D, Walters M. POSSUM: A scoring system for surgical audit. Br J Surg 1991;78:355-60. Crossref

6. Copeland G. The POSSUM system of surgical audit. Arch Surg 2002;137:15-9. Crossref

7. Khuri S, Daley J, Henderson W, et al. The National VA Surgical Risk Study: risk adjustment for the comparative assessment of the quality of surgical care. J Am Coll Surg 1995;180:519-31.

8. Daley J, Khuri SF, Henderson W, et al. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs surgical risk study. J Am Coll Surg 1997;185:328-40. Crossref

9. Daley J. Validating risk-adjusted surgical outcomes: site visit assessment of process and structure. J Am Coll Surg 1997;185:341-51. Crossref

10. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume One: July 2008-June 2009. Hospital Authority, Hong Kong SAR Government; 2010.

11. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Two: July 2009-June 2010. Hospital Authority, Hong Kong SAR Government; 2011.

12. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Three: July 2010-June 2011. Hospital Authority, Hong Kong SAR Government; 2012.

13. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Four: July 2011-June 2012. Hospital Authority, Hong Kong SAR Government; 2013.

14. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Five: July 2012-June 2013. Hospital Authority, Hong Kong SAR Government; 2014.

15. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Six: July 2013-June 2014. Hospital Authority, Hong Kong SAR Government; 2015.

16. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Seven: July 2014-June 2015. Hospital Authority, Hong Kong SAR Government; 2016.

17. Surgical Outcomes Monitoring & Improvement Program (SOMIP) report. Volume Eight: July 2015-June 2016. Hospital Authority, Hong Kong SAR Government; 2017.

18. Ash A, Schwartz M. Evaluating the performance of risk-adjustment methods: dichotomous variables. In: Iezzoni L, editor. Risk adjustment for measuring health care outcomes. Ann Arbor, MI: Health Administration Press; 1994: 313-46.

19. Hall BL, Hamilton BH, Richards K, Bilimoria KY, Cohen ME, Ko CY. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: An evaluation of all participating hospitals. Ann Surg 2009;250:363-76. Crossref

20. Cohen ME, Liu Y, Ko CY, Hall BL. Improved surgical outcomes for ACS NSQIP hospitals over time: evaluation of hospital cohorts with up to 8 years of participation. Ann Surg 2016;263:267-73. Crossref

Organ transplantation offers the best clinical management for patients with end-stage organ failure. Since the first successful kidney transplant in 1954,1 a number of studies have confirmed that organ transplantation provides the best outcomes for survival,2 quality of life,3 and cost-effectiveness.4 With advances in surgical techniques and immunosuppressive drugs, organ donation has also progressively improved over the decades. However, the low rate of deceased organ donation is a universal problem. Different countries have adopted different policy schemes to encourage deceased organ donation but the donation rates vary.

Currently, Hong Kong adopts an opt-in legislative system of organ donation, where an individual who is willing to donate their organs after death is required to carry a signed donor card or register online at the Central Organ Donation Register.5 However, the registration rate among the general population is low and was about 11% in 2015.6 The actual number of deceased donors in the same year was only 5.8 per million population, which is much lower than in other developed areas: for example, Spain has a deceased donor rate of 34 per million population, which is the world’s highest.7 As a result, organ demand consistently outweighs supply, leading to a long waiting list.8

The World Health Organization is urging every country to advance their organ procurement programme.9 Two legislative systems have been proposed to increase the donation rate: an opt-out system and an opt-in system with organ allocation priority (allocation priority system).10 An opt-out system presumes an individual is a potential deceased donor unless refusal has been expressed by ‘opting out’ of the donor pool. This system can simplify registration procedures and is convenient for those willing to donate organs. An opt-out system, however, does not specifically indicate willingness.11 On the basis of empirical evidence from several European countries, implementing an opt-out system can successfully raise awareness and willingness for deceased organ donation12 and actual donation rates.11 13 14 15

An opt-in system with allocation priority is another policy approach, in which individuals who have registered as a deceased donor will gain priority points on the transplant waiting list. A priority point is a reward for those who are willing to donate an organ and who in turn gain preferential status for receiving a donor organ if required.16 This system can motivate the public to register as donors by providing them with a higher chance of extending their own lives and address the perceived unfairness of ‘free-riders’ who are willing to receive an organ but unwilling to donate. Israel adopted this policy approach in the Organ Transplant Act of 2008 and illustrated its effectiveness by the number of signed donor cards and actual donation rates.10 16 17

In light of the low deceased organ donation rate in Hong Kong, the government is currently reviewing its donation policy and is considering replacing the current opt-in system by an opt-out system. However, introducing an opt-out system without public support may actually reduce the donation rate,18 as it did in Brazil and Singapore.6 Similarly, ethical issues of the allocation priority system may fuel public opposition. On the one hand, it promotes a fair concept of reciprocity.16 19 On the other hand, there is a perception that organ allocation should be based not on an individual’s willingness to donate, but solely on their medical needs.16 19 It is difficult to predict the effectiveness of proposed legislative systems in Hong Kong on the basis of experience elsewhere, and local analysis of these systems is limited. Only one local study has examined the willingness to donate a kidney under an opt-out system but the feasibility of allocation priority was not included.20 In addition, socio-economic and demographic determinants may influence willingness towards deceased organ donation. A Malaysian study suggested that such determinants may differ under different legislative systems.21 Interestingly, political viewpoint is also correlated with willingness when a new legislative system is imposed.12 22 It is thus important to explore these determinants when considering an opt-out system or allocation priority system.

This study aimed to explore whether a proposed opt-out system or opt-in allocation priority system would increase public willingness in Hong Kong towards deceased organ donation, and to examine the association of socio-economic, demographic, and political determinants with the willingness of deceased organ donation under different proposed policies.

A cross-sectional telephone survey with a structured questionnaire was conducted. The target population was Hong Kong permanent residents who were aged ≥18 years, able to register as a deceased organ donor, and able to speak Cantonese. On the basis of the ‘10k rule of thumb’,23 the minimum sample size required to test the association between the willingness of organ donation and the seven socio-demographic variables in this study was about 200. Assuming a 0.05 type 1 error, the sample size would be able to detect 15% and 20% significant differences in the proportions of persons shifting from ‘unwilling’ to ‘willing’ to donate after proposing the allocation priority system and opt-out system, respectively, both with at least 80% power. With the assumption that 12% of landline telephone numbers are valid and there would be a 30% response rate, a sample of about 5600 telephone numbers was needed. A computerised random list of 5800 eight-digit residential telephone numbers (starting with ‘2’ and ‘3’, the current telephone number system in Hong Kong) was generated.

Telephone interviews were held from 18:00 to 20:30 on every weekday and from 11:00 to 18:00 on every weekend to ensure coverage of different demographics. The actual proportions of calls made on weekdays and at weekends were 41.3% and 58.7%, respectively. Interviews were conducted by the first two authors and each took about 5 to 7 minutes, including an introduction explaining the interview and obtaining verbal consent. A maximum of three calls was made at different times on different days before a telephone number was considered invalid. If more than one family member in a household was eligible, the person whose next birthday was closest to the interview date was invited to participate.

The questionnaire was developed from a literature review and comprised 14 items in four sections: (1) current opt-in system; (2) proposed opt-out system; (3) proposed opt-in allocation priority system; (4) background information: socio-economic and demographic characteristics and political views. Reasons underlying the willingness under each legislative system were explored with open-ended questions. On the basis of a pilot study of the questionnaire’s feasibility and wording, involving 10 members of the general public, descriptions of the two proposed systems were refined. Double-entry of data and data cleaning were conducted by the authors. Ethics approval was granted by the Survey and Behavioural Research Ethics Committee of The Chinese University of Hong Kong.

All statistical analyses were performed with IBM SPSS Statistics 24 (IBM Corp, Armonk [NY], USA). A descriptive summary of characteristics of respondents compared with those of the targeted general population24 was prepared. Two-sided McNemar’s tests were used to compare willingness rates between the current opt-in system and the opt-out system, as well as between the current opt-in system and the allocation priority system. Univariate logistic regression analyses were performed to explore associations between each independent variable (demographics, socio-economic level, and political view) and the dependent variable (willingness of deceased organ donation) under the proposed opt-out and allocation priority systems, separately. Significant independent variables were further tested by multiple logistic regression analysis using the forced entry method. P<0.05 was considered statistically significant. A qualitative content analysis was also conducted by the first author to identify meaningful units emerging from the open-ended questions for underlying reasons for willingness to donate. The units were coded into categories based on the findings and suggested by the literature review,6 16 17 25 and were checked independently by the second author.

Telephone interviews took place between 2 August and 18 October 2016 in Hong Kong. A total of 5800 households were contacted and 203 respondents completed the questionnaire (response rate, 19.7%) (Fig).

The majority of the respondents were female (60.9% of 202), were aged 18 to 30 years (43.2% of 199), had a post-secondary education (41.0% of 200), were employed (54.0% of 202), were never married (52.3% of 197), and had a monthly household income of more than HK$40 000 (37.6% of 189). Compared with the general population, our study sample contained smaller proportions who were male, older than 40 years, educated to secondary level, or married, or who had a household income of <HK$20 000. (Table 1).

Under the current opt-in system, a majority of respondents were willing to donate organs after death (64.5%), whereas some were unwilling (11.8%) and others were unsure (23.6%) [Table 2]. Among those willing to donate, only 29.0% had registered as deceased donors.

After the opt-out system was proposed, 37.0% of respondents who were unsure about donation or had been unwilling to donate previously, reported a willingness to stay in the donor pool. However, 15.8% of registered donors and 28.2% of respondents who had originally expressed a willingness to donate were reluctant to stay in the donor pool. Overall, the willingness rate decreased from 64.5% under the current opt-in system to 60.1% under the proposed opt-out system, although the reduction was not statistically significant. In contrast, combining the opt-in system with allocation priority motivated 64.8% of respondents originally unwilling to donate to instead express a willingness. Only 11.4% of those originally willing to donate were demotivated. This resulted in a statistically significant increase in the overall willingness rate, from 64.5% under the current opt-in system to 73.4% under the proposed allocation priority system (P=0.018) [Table 2].

Under the allocation priority system, respondents who changed their stance from unwilling to willing to donate (n=35) did so largely because they perceived the system as a “fair or reciprocal” method of organ allocation (57.1%). They believed they “had a greater chance of receiving an organ transplant if registered” (54.3%). At the same time, some respondents who had originally expressed a willingness to donate were reluctant to do so under the priority system (n=17). They felt that the preferential status was useless, as “the waiting list would be still very long” (52.9%). Some also mentioned that it was “unethical” if criteria other than medical condition were used for organ allocation (47.1%) [Table 3].

After the opt-out system was explained, those who had been unsure about donor registration or unwilling to donate under the opt-in system (n=30) were willing to stay in the donor pool because the opt-out system offered them “convenience for indicating willingness” (23.3%). Moreover, some believed that remaining in the donor pool was a “civic duty” (20.0%). Conversely, many respondents who had been willing to donate under the opt-in system chose to opt out (n=39), as they perceived that under the system they were “forced to donate” (69.2%). “Distrust of local government” was another reason (30.8%) [Table 4].

Associations between variables and willingness were first tested by univariate logistic regression for each legislative system (Table 5). Under the opt-out system, age, educational attainment, monthly household income, marital status, and political view were significantly associated with willingness. Under the allocation priority system, age, educational attainment, monthly household income, and marital status were the significant predictors. Sex and occupation were not significantly associated with a willingness to donate under either system.

These significant predictors were further analysed by multiple logistic regression. Under an opt-out system, respondents who were married (adjusted odds ratio [AOR]=0.423, 95% confidence interval [CI]=0.205-0.871) or earned a monthly household income of <HK$20 000 (AOR=0.366, 95% CI=0.162-0.827) or HK$20 000-39 999 (AOR=0.447, 95% CI=0.207-0.962) had a lower willingness to donate, whereas those who trusted local government had a higher willingness (AOR=2.590, 95% CI=1.023-6.554). Under an allocation priority system, lower willingness to donate was associated with age over 60 years (AOR=0.168, 95% CI=0.029-0.960), primary education level or no schooling (AOR=0.253, 95% CI=0.077-0.829), and monthly household income of <HK$20 000 (AOR=0.230, 95% CI=0.080-0.622).

This study provides the first analysis of attitudes towards different policies for deceased organ donation and related determinants. We found that an allocation priority system would significantly motivate respondents to donate their organs, similar to the findings of an Israeli public telephone survey.17 The major underlying reason of “fairness or reciprocity” is aligned with the concept of justice,19 as organs are a scarce societal resource with a demand that heavily outweighs the supply. The positive effect of a proposed priority incentive on respondents’ willingness to donate may be related to possible organ scarcity in the market with the extremely low donation rates in both Israel and Hong Kong.7 In addition, the concept of reciprocity might be derived from a moral duty of mutual aid.26 Similar to many countries that adopt the priority system, especially Singapore, Hong Kong treats moral duty as a legislative foundation. Respondents might have agreed that those who refuse to donate their organs (free-riders) should not receive organs ahead of those who are willing to donate.17 Furthermore, respondents ,may also have been motivated by priority incentives, providing them with a potential chance to extend their life. This outcome is unsurprising, as it is the key feature of this allocation priority system.16

With the increasing demand for organ transplantation, the Hong Kong government has explored the feasibility of an opt-out system. It is worth noting that a proposed opt-out system caused a reduction in the willingness for deceased organ donation, although not to a significant degree. This finding contradicts that suggested by a recent Hong Kong study on kidney donation,20 which claimed that the willingness to donate would rise significantly under an opt-out system. This inconsistency may be attributable to an assumption made by that study, that those willing to donate organs under the current opt-in system would remain willing under an opt-out system. Yet, our study found that a large number of respondents who were originally willing to donate changed to being unwilling to stay in the donor pool. The switch was because many initially willing respondents perceived that under an opt-out system, they were “being forced” to donate. Nonetheless, the opt-out system has successfully induced a willingness to donate in many European countries12 that also advocate personal liberties. According to a European study, a societal environment was a prerequisite for the government to justify an opt-out system that would limit citizens’ liberties.27 First, the opt-out system can be imposed only when there are no less restrictive alternatives. Yet, alternatives do exist, such as the allocation priority system. Second, as in the present study, government popularity seems to be another prerequisite condition that was lacking. Distrust of the local government led some respondents to opt out.

Moreover, our study identifies determinants associated with a willingness towards deceased organ donation. Respondents with a lower education level and older age were less likely than others to donate organs under an allocation priority system. These significant factors might arise from a stronger traditional belief among the elderly population of keeping a body intact after death,21 as well as a lower awareness of organ donation among those with less education.28 The present study also reveals that those who were married were more reluctant than never-married people to donate under the opt-out system, because a married person might need a partner’s consent before making a decision about donating organs. Echoed by Malaysian and European studies, a trust in government was associated with a higher willingness towards deceased organ donation under the opt-out system.12 22

Lower monthly household income was associated with lower willingness to donate organs under both systems. This association may be because those with a higher income are more likely to promote a supportive attitude towards organ donation.29 Although the significance of each determinant varied between the opt-out system and the allocation priority system, strengths and directions of the associations were similar across both. More importantly, these significant determinants were also significant under the current opt-in system.30 Thus, regardless of the legislative system imposed, determinants associated with willingness to donate appear the same.

This study provides preliminary evidence of the potential effectiveness of different legislative systems. In particular, the study responds to the recent public controversy over the possible introduction of an opt-out system in Hong Kong,31 and may help policymakers anticipate public opposition to such a system. The government should first create a supportive societal environment and gain public trust before its implementation.

At the same time, this study provides evidence to support adding allocation priority to the current opt-in system in Hong Kong. The findings offer policymakers new insight into alternative legislative systems other than the opt-out system. Further evaluation of the priority incentive or other policy instruments is suggested so that policymakers can identify the best alternative.

The design of an administrative procedure to motivate and facilitate registering behaviour should also be considered. Similar to other local studies,6 30 this study found that the rate of registration to donate was quite low among those currently willing. The main reasons may be laziness and lack of knowledge about the registration procedure.6 Thus, an individual’s attitude towards donation is not necessarily aligned with registering behaviour.12 32 In other words, combining the opt-in system with allocation priority may not necessarily result in a higher registration and donation rate in practice. Policymakers should consider measures that will simplify the registration procedure. For example, in the United States and the United Kingdom, driving license applicants are invited to register as deceased organ donors.33

Another recommendation stemming from this study is the development of targeted promotion strategies when a new legislative system is introduced. With an understanding of determinants of willingness to donate, the Organ Donation Promotion Charter can target those who are less willing to donate. By increasing knowledge and alleviating concerns about procedures involved under the new legislative system, public willingness is expected to increase.12 Promotional campaigns should also help build public trust in the government for a smooth implementation of the new system.

The strength of this study is the use of random sampling for respondent recruitment. A random-digit dialling method was used such that unlisted numbers were also contacted. Each residential telephone number, therefore, had an equal selection probability. With a 95% residential fixed-line penetration rate,34 the sampling frame included most of the Hong Kong general population.

This study has several limitations. There may have been selection bias (selective timing of telephone calls) and self-selection bias (non-response after receiving phone calls). Our study is not representative of the general population, as it has fewer respondents who were male, of older age, educated to secondary level, of lower socio-economic status, and married. In addition, without standardised protocols, information bias may have arisen from recording and classifying responses from the open-ended questions that asked for underlying reasons for change in willingness. Examination of determinants of the willingness to donate was also limited by the small sample size. Subsequent surveys with a larger sample are recommended to investigate socio-demographic variables as well as other possible factors, such as chronic illness requiring an organ transplant in respondents and their relatives or friends.

This study examined the impact of a proposed opt-in system with organ allocation priority and an opt-out system on willingness towards deceased organ donation among the Hong Kong general population. An allocation priority system could induce willingness to donate. At the same time, the study provides discouraging evidence for the effectiveness of an opt-out system. These findings have implications for policy-making and targeted education. More research is needed to study alternative legislative systems to solve the crisis of organ shortage in Hong Kong.

We thank Ms Yuen-Fan Tong for her support and experience in the development of the questionnaire.

The authors have no conflicts of interest to disclose.

1. Matesanz R, Dominguez-Gil B. Strategies to optimize deceased organ donation. Transplant Rev 2007;21:177-88. Crossref

2. Wolfe RA, Ashby VB, Milford EL, et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med 1999;341:1725-30. Crossref

3. Keown P. Improving quality of life—the new target for transplantation. Transplantation 2001;72(12 Suppl):S67-74.

4. Winkelmayer WC, Weinstein MC, Mittleman MA, Glynn RJ, Pliskin JS. Health economic evaluations: the special case of end-stage renal disease treatment. Med Decis Making 2002;22:417-30.

5. Department of Health, Hong Kong SAR Government. Know more about centralised organ donation register. 2013. Available from: https://www.organdonation.gov.hk/eng/knowmore.html. Accessed 2 Nov 2016.

6. Hong Kong Ides Centre. A preliminary study on deceased organ donation in Hong Kong. 2015. Available from: http://www.ideascentre.hk/wordpress/wp-content/uploads/2009/02/Final-Report-for-Organ-Donation-TC.pdf. Accessed 2 Nov 2016.

7. International Registry in Organ Donation and Transplantation. 2015. Available from: http://www.irodat.org/. Accessed 2 Nov 2016.

8. Department of Health, Hong Kong SAR Government. Statistics (milestones of Hong Kong organ transplantation). 2016. Available from: https://www.organdonation.gov.hk/eng/statistics.html. Accessed 2 Nov 2016.

9. Rodriguez-Arias D, Wright L, Paredes D. Success factors and ethical challenges of the Spanish Model of organ donation. Lancet 2010;376:1109-12. Crossref

10. Zúñiga-Fajuri A. Increasing organ donation by presumed consent and allocation priority: Chile. Bull World Health Organ 2015;93:199-202. Crossref

11. Johnson EJ, Goldstein D. Medicine. Do defaults save lives? Science 2003;302:1338-9. Crossref

12. Mossialos E, Costa-Font J, Rudisill C. Does organ donation legislation affect individuals’ willingness to donate their own or their relative’s organs? Evidence from European Union survey data. BMC Health Serv Res 2008;8:48. Crossref

13. Rithalia A, McDaid C, Suekarran S, Myers L, Sowden A. Impact of presumed consent for organ donation on donation rates: a systematic review. BMJ 2009;338:a3162. Crossref

14. Abadie A, Gay S. The impact of presumed consent legislation on cadaveric organ donation: a cross-country study. J Health Econ 2006;25:599-620. Crossref

15. Gimbel RW, Strosberg MA, Lehrman SE, Gefenas E, Taft F. Presumed consent and other predictors of cadaveric organ donation in Europe. Prog Transplant 2003;13:17-23. Crossref

16. Cronin AJ. Points mean prizes: priority points, preferential status and directed organ donation in Israel. Isr J Health Policy Res 2014;3:8. Crossref

17. Siegal G. Making the case for directed organ donation to registered donors in Israel. Isr J Health Policy Res 2014;3:1. Crossref

18. Institute of Medicine. Presumed consent. In: Childress FJ, Liverman TC, editors. Organ Donation: Opportunities for Action. Washington, DC: The National Academies Press; 2006: 214.

19. Chandler JA. Priority systems in the allocation of organs for transplant: should we reward those who have previously agreed to donate. Health Law J 2005;13:99-138.

20. Chan TK, Cowling BJ, Tipoe GL. A public opinion survey: is presumed consent the answer to kidney shortage in Hong Kong? BMJ Open 2013;3.pii:e002013. Crossref

21. Tumin M, Tafran K, Mutalib MA, et al. Demographic and socioeconomic factors influencing public attitudes toward a presumed consent system for organ donation without and with a priority allocation scheme. Medicine (Baltimore) 2015;94:e1713. Crossref

22. Tumin M, Noh A, Jajri I, Chong CS, Manikam R, Abdullah N. Factors that hinder organ donation: religio-cultural or lack of information and trust. Exp Clin Transplant 2013;11:207-10. Crossref

23. Troutt MD. Regression, 10k rule of thumb for. In: Kotz S, Campbell RB, Balakrishnan N, Vidakovic B, Johnson NL, editors. Encyclopedia of Statistical Sciences. John Wiley & Sons, Inc; 2004:7098.

24. Census and Statistics Department, Hong Kong SAR Government. By-census results. 2016. Available from: http://www.bycensus2016.gov.hk/en/bc-index.html. Accessed 2 Nov 2016.

25. Lavee J, Brock DW. Prioritizing registered donors in organ allocation: an ethical appraisal of the Israeli organ transplant law. Curr Opin Crit Care 2012;18:707-11. Crossref

26. Peters DA. A Unified approach to organ donor recruitment, organ procurement, and distribution. J Law Health 1988-1989;3:157-87.

27. Verheijde JL, Rady MY, McGregor JL, Friederich-Murray C. Enforcement of presumed-consent policy and willingness to donate organs as identified in the European Union Survey: the role of legislation in reinforcing ideology in pluralistic societies. Health Policy 2009;90:26-31. Crossref

28. Riyanti S, Hatta M, Norhafizah S, et al. Organ donation by sociodemographic characteristics in Malaysia. Asian Soc Sci 2014;10:265-72.

29. Wong LP. Knowledge, attitudes, practices and behaviors regarding deceased organ donation and transplantation in Malaysia’s multi-ethnic society: a baseline study. Clin Transplant 2011;25:E22-31. Crossref

30. Surveillance and Epidemiology Branch, Centre for Health Protection, Department of Health, Hong Kong SAR Government. Behavioural risk factor survey (April 2011): main report. 2012. Available from: http://www.chp.gov.hk/files/pdf/brfs_2011apr_en.pdf. Accessed 2 Nov 2016.

31. Tsang E. Hong Kong to discuss organ donation opt-out scheme following death of girl awaiting double lung transplant. South China Morning Post. 2015 Oct 9.

32. Ugur ZB. Does presumed consent save lives? Evidence from Europe. Health Econ 2015;24:1560-72. Crossref

33. Buckley TA. The shortage of solid organs for transplantation in Hong Kong: part of a worldwide problem. Hong Kong Med J 2000;6:399-408.

34. Office of the Communications Authority, Hong Kong SAR Government. Key communications statistics. 2016. Available from: http://www.ofca.gov.hk/en/media_focus/data_statistics/key_stat/. Accessed 2 Nov 2016.

This article provides an up-to-date overview of breast cancer mammography screening and briefly discusses its history, controversies, current guidelines, practices across Asia, and future directions. An emphasis is made on shared decision-making—instead of giving just a ‘yes’ or ‘no’ answer to patients, the focus should be on providing sufficient information about the pros and cons of screening to help women make a personal, informed choice. Frontline experts, including breast surgeons, oncologists, breast radiologists, and their representative professional associations should all participate in guideline panels, with the goal of improving cancer detection, reducing mortality, and improving patient outcome.

This article provides an up-to-date overview of breast cancer mammography screening and briefly discusses its history, controversies, current guidelines, practices across Asia, and future directions. An emphasis is made on shared decision-making—instead of giving just a ‘yes’ or ‘no’ answer to patients, the focus should be on providing sufficient information about the pros and cons of screening to help women make a personal, informed choice.

The goal of mammographic screening (and other breast-cancer screening tests) is to detect breast cancer earlier than it would otherwise manifest clinically, when it is less likely to have spread. Data clearly show that detection of breast cancers at smaller sizes and lower (earlier) stages is associated with better patient outcomes, lower morbidity, and reduced breast cancer deaths.1 Reduced morbidity is likely to be related to feasibility of breast conservation and hence less extensive surgery, fewer associated complications such as lymphoedema, less chemotherapy, and hence fewer adverse effects.2 Other benefits of diagnosing screen-detected cancers at an earlier stage also include a lower cost of treatment and consequent reduced financial burden on health care resources.3

The Table summarises the mammography guidelines from selected nations.4 5 In common, all organisations emphasise that the benefits of screening outweigh the harm at all ages.3 6 They all endorse informed decision-making and the importance of informing women about both benefits and limitations of screening. However, there remain legitimate concerns about guideline differences, including the complexity of the guidelines; weak adherence to creating opportunities for informed decision-making; unreadiness of referring clinicians to discuss benefits, limitations, and harm associated with screening; and the lack of reminder systems, which results in weaker adherence to recommended screening intervals. Despite these concerns, it is widely accepted that high adherence to even the least aggressive guidelines will save more lives than the current weak adherence to regular screening programmes.4

Randomised controlled trials (RCTs) have been the gold standard for proving that early detection with mammography decreases mortality from breast cancer. Since the very first screening RCT performed in New York in the 1960s, there have been eight prospective RCTs and numerous subsequent meta-analyses published. Most well-executed RCTs demonstrated a 20% to 30% decrease in mortality from breast cancer when women were invited for screening. These results laid a solid foundation for population-based screening programmes worldwide.1 7 8

Subsequent studies that generated data from population-based screening programmes have provided further evidence of the benefits of screening mammography. The true benefit reported (in terms of mortality reduction) ranged from 38% to 49%, even higher than that shown by RCTs. This difference demonstrates that service screening studies measure the direct effect of screening on women who actually underwent mammography, and not just those who were invited to undergo mammography (as opposed to the methodology of RCTs). Service screening studies also tend to measure the effect of more recent screening practices that have benefited from improved mammography technology, better breast positioning techniques, and improved interpretive skills.1 9

One exception to the RCTs that reported unfavourable results of mammographic screening was the Canadian National Breast Screening Study (CNBSS). It was conducted between 1980 and 1985, and was divided into two parts. The first CNBSS included approximately 50 000 volunteer women aged 40 to 49 years, and determined the mortality benefit in the experimental group, who were assigned to annual screening mammography plus clinical breast examination (CBE) versus the control group who received usual care.10 The second CNBSS had almost 40 000 volunteer women aged 50 to 59 years, and compared the benefit of annual mammography plus CBE with that of yearly CBE alone.11

From the time the results were first published in 1992 and again after follow-up in 2000, 2002, and 2014, the CNBSS has been controversial, because it is the only RCT to have reported no decrease in mortality associated with an invitation to screening. The study also claimed a 22% overdiagnosis of screen-detected invasive cancer, increasing to up to 35% when cases of ductal carcinoma in situ (DCIS) were included.12 13 14 However, the credibility and scientific value of the CNBSS study have been repeatedly questioned in peer-reviewed publications.15 16 17 18 19 Most criticisms of this study are related to vulnerabilities and shortcomings in its execution, including flaws in the randomisation process, lack of statistical power, non-generalisable results, poor quality imaging, suboptimal mammographic image acquisition and interpretation by untrained personnel, and inconsistent thresholds for interpretation.

The flaws in the randomisation process principally arose from three areas. First, unlike all other RCTs, potential participants in the Canadian trials initially underwent a careful physical examination. Second, women with positive findings on physical examination, including palpable lumps, skin or nipple retraction, and even palpable axillary adenopathy, were not excluded from this ‘screening’ trial.18 Finally, randomisation was unblinded and decentralised. Because almost 80% of women with advanced palpable cancers were assigned to the screening arm in the first round of the study, there has been speculation that concerned clinicians did not follow the randomisation process, but rather assigned some symptomatic women to the study group so that they would undergo mammography.19 Whether the imbalance was due to intentional tampering or occurred by chance alone, the net effect was the same—namely, a failure to produce two equal cohorts of patients for comparison.

The CNBSS was also criticised at the time of the trial for poor quality mammography, even compared with mammographic imaging of that era.15 20 To reduce radiation dose, mammography for the trial was performed without the benefit of scatter-reducing grids despite their routine use and availability. Standard imaging for much of the trial used a straight lateral view, not a mediolateral-oblique view, which images more tissue. The combination of poor quality imaging and the investigators’ resistance to taking corrective action led two advisors’ resignation in protest. In addition, technologists who participated in the trial received no special training in performing mammography. Radiologists new to mammography also received no training in interpretation.18 There was also a lack of immediate follow-up after recommendations for biopsy had been made. Overall, about 25% of the recommended biopsies were ultimately not performed.18

The CNBSS trials are an excellent example of the need to carefully consider all facets of a large-scale screening trial before accepting its results as scientifically valid. The numerous design and execution flaws described above explain in large part why the results of the CNBSS are dramatically different from those of all other RCTs. Ultimately, on the basis of the methodology of the CNBSS, the World Health Organization excluded those results when analysing the breast-screening data in the International Agency for Research on Cancer report.21

The greatest debate on the value of breast screening arose after the publication of a highly controversial but frequently quoted meta-analysis by Gotzsche (a medical statistician and director of the Nordic Cochrane Centre) and Olsen in The Lancet in 2000. Their study concluded that there was no benefit of mortality reduction by screening, after discarding six of eight RCTs because they deemed the randomisation to be “inadequate”. The only two RCTs included in their analysis showed no benefit, including the Malmo trial and the notorious CNBSS.7 22

Gotzsche and Olsen’s critique and methodology have caused much controversy and, in turn, have been criticised heavily by leading expert breast imagers, public health clinicians, and professional bodies such as the Society of Breast Imaging.7 8 23 24 25 26 27 Gotzsche and Olsen’s use of quoted figures from cancer registries rather than actual patient data, their selective approach to studies, and in particular the ignoring of the flaws of the CNBSS, have received the harshest criticism. Many experts have commented that Gotzsche and Olsen overstated the limitations of most of the well-executed RCTs, thereby reflecting a “context-free” application of guidelines in a way that did not address the real issues relevant to the effectiveness of mammographic screening. Moreover, Gotzsche and Olsen’s recommendation to abandon screening altogether has hampered collaborative efforts to improve breast cancer detection and control.27

In February 2014, the Swiss Medical Board attempted to overturn the widespread practice of mammography screening in Switzerland by stating that new systematic mammography screening programmes should not be introduced, irrespective of women’s age, and recommended that existing programmes should be discontinued. Their main argument was that the absolute risk reduction in breast cancer mortality was low and that the adverse consequences of screening (false-positive test results, overdiagnosis, overtreatment, and high costs and expense of follow-up tests and procedures) were substantial.28 29

The Swiss Medical Board’s attempt initiated a new phase of heated arguments and debate about the benefits of screening. Expert breast cancer clinicians in both the United States and Europe (including leading cancer associations in Switzerland) rejected their report. One criticism was that the Swiss Medical Board relied heavily on the controversial work by Gotzsche and Olsen and again quoted data from the flawed CNBSS. Another criticism that attracted great attention was the questionable “expert panels” of the board: they included a medical ethicist, a clinical epidemiologist, a clinical pharmacologist, an oncology surgeon, a nurse scientist, a lawyer, and a health economist. Frontline breast imagers, with expertise in diagnosing breast diseases, were excluded from the review panels because of a “conflict of interest”.28 29

The Swiss Medical Board did not adequately consider the fact that assessment of the balance between benefit and harm involves a value judgement that each woman should make only after she is fully informed about the strengths and weaknesses of screening mammography. They also disregarded the extensive literature in support of screening mammography (RCTs and population service screening studies), making their attempt at stopping national mammography screening unjustified.

Commonly mentioned potential harms of screening include false-positive mammograms, recall for additional imaging, a false-positive biopsy, missed breast cancer, radiation dose, patient anxiety, and, above all, overdiagnosis.

Overdiagnosis is defined as the detection (and subsequent actions taken) of a cancer by screening that would not have progressed to become symptomatic in a woman’s lifetime.1 The estimation of overdiagnosis is complex, highly debated, and very difficult to measure.3 Reported figures range widely, from 0% to 50%, vary greatly in terms of methodological rigour, and testify to the inexact nature of most mathematical models.30 31 32 33 34 When appropriate adjustments for temporal trends, risk factors, and lead time are considered, the level of overdiagnosis should be low, within the range of 0% to 10%.32 Importantly, a recent study of over 5 million women (aged 50-64 years) screened by the United Kingdom’s National Health Service showed that there was a significant negative association between the detection of DCIS at screening and invasive interval cancers. In that study, Duffy and colleagues analysed the data from four consecutive screen years and the 36-month outcome after each relevant screen. For every three screen-detected cases of DCIS, there was one less interval case of invasive cancer over the next 3 years. They agreed that the policy on detection and treatment of DCIS is worthwhile and can prevent subsequent invasive cancers.35

The effect of screening on heightening a patient’s anxiety has also been long questioned by critics, but the magnitude of the effect may have been over-exaggerated. In a survey of over 1200 women with a 6-question anxiety scale to understand the short-term and long-term impact of a recall examination, women involved in the digital mammographic imaging screening trial demonstrated only a transient, limited increase in anxiety after a false-positive mammogram compared with those with a negative mammogram, and there was no difference between the two groups’ intention to undergo mammography again in the subsequent 2 years.36 Schwartz et al reported that 96% of American women who received a false-positive mammography report were glad that they underwent the test and remained supportive of screening.37 Most women agreed that the anxiety, inconvenience, and the few image-guided needle biopsies using local anaesthesia associated with a recall from screening, were minor compared with dying of breast cancer.38

To summarise, papers citing a high rate of overdiagnosis in screening (in the magnitude of 20% or higher) and claiming that false-positives are a significant cause of patient anxiety are believed by most experts to be overstating the case.

Although it is important to discuss all aspects of screening asymptomatic women (including potential harm), the harm of not attending screening is underestimated and not discussed. For instance, women who do not attend screening have significantly larger tumours, a higher stage at diagnosis, poorer overall and disease-specific survival, and higher costs of treatment.39 It has been estimated that the cost of treating advanced metastatic breast cancer exceeds USD 250 000 per patient, and the average cost of treating advanced cancer in the first year after diagnosis is almost double that of early cancers, mainly owing to the difference in costs of chemotherapy.3 40 The cost of treatment and lost productivity each year will far exceed the cost of annual screening and, additionally, do not include the indirect value of the lives saved (as a productive member of workforce).1

The incidence of breast cancer continues to increase worldwide. It remains highest in the United States and Europe, but has been increasing substantially in Asian countries over the past three decades.41 Studies that compare invasive breast cancer data from Asia with those from the United States over a 20-year period have shown that female breast cancer incidence among Asian and Western populations is more similar than expected.42 The incidence of female breast cancer in China will continue to rise, and is expected to exceed 100 per 100 000 women by 2021, giving a total of 2.5 million cases.43

According to GLOBOCAN 2012 of the International Agency for Research on Cancer, the specialised cancer research agency of the World Health Organization, almost a quarter (24%) of all breast cancers were diagnosed within the Asia-Pacific region, with the greatest number occurring in China (46%).44 The age-standardised incidence rate was highest among Taiwanese (65.9 per 100 000), followed by Singaporeans, South Koreans, and Japanese.44 In a multiracial country such as Singapore, Chinese women have been noted to have a significantly higher risk of developing breast cancer than Malays and Indians.45

The disease burden in Hong Kong is no different. Locally, the age-standardised incidence rate was 58.8 per 100 000 in 2015, with over 3900 new cases per year.46 A study of the local trend in female breast cancer incidence from 1973 to 1999 by the University of Hong Kong showed a significant yearly increase of an average of 3.6%; the increase was most marked and continued to accelerate in the younger age-groups. It was speculated that such trend changes were related to Westernisation of lifestyle.47 All these data indicate that the disease burden in Hong Kong is increasing and comparable to that of all other civilised Asian countries and cities.

Breast screening services in Asian countries and cities are highly variable: some have advanced nationwide screening programmes and others have less developed programmes.48 South Korea and Taiwan are both well recognised for their experience in running such programmes, the former having the highest intake rate and the latter being the most well-structured.

South Korea places a very strong emphasis on screening for cancer control in general. Its national health service offers mammography and CBE every 2 years to women aged 40 or older, and at no cost to the 50% of people with the lowest incomes. Their programme is popular and widely accepted by the general public, and achieved an uptake of as high as 66% in 2014. Benefits of downstaging from screening were also observed. However, South Korea encountered a problem of potential overdiagnosis, with a noticeably higher false-positive rate when compared with other places.

Taiwan’s health authorities have been recognised for rolling-out well-organised and well-resourced screening programmes, with good support from a local randomised controlled trial showing a reduction in mortality by 40% with mammography screening.49 Since 2004, their health service has provided free breast screening to women aged 50 to 69 years, expanded in 2010 to those aged 40 to 49 years. By 2015, about 40% of the target population participated in screening. It is believed that the cause of the suboptimal participation rate was not due to capacity or outreach, but rather the Taiwanese public’s values and attitude. Nonetheless, with more resources being directed to public education and motivation, Taiwan’s health authorities are pushing their goal to 60% by 2018.

The experience of screening programmes in Singapore and Japan is more equivocal. Despite having sufficient scientific evidence to support their role in reducing mortality and reducing invasive cancer incidence, the participation rate has remained lower than expected, mostly owing to cultural barriers and paradigms, or a lack of central governing. Singapore established its national, population-wide screening programme (BreastScreen Singapore) in 2002 and now covers women aged 40 to 69 years. The participation rate has been noted to plateau at 40% since 2010, short of the target of 70%. The health promotion board believes that apart from cultural issues, costs (as screening is paid by an individual’s medical insurance account) constitute the greatest barrier to uptake.

The study of population-based screening in Japan has been complex, with scattered data owing to the lack of a single national organisation for monitoring. The participation rate remains lower than in other comparable Asian countries in the past century, again likely because of cultural paradigms. Despite these barriers, in the past decade, Japanese health officials have started designing their own methods and protocols for screening, particularly targeting the higher incidence of cancer among younger women (aged 40-49 years) and the large proportion of patients with dense breasts. After the launch of government-funded screening programmes, a clinical trial that started in 2007 (Japan Strategic Anti-cancer Randomised Trial, J-START) of over 70 000 women undergoing adjunctive ultrasonography to supplement mammography for screening showed an increased sensitivity and detection rate for early preclinical cancers.41

In China, there is no nationwide screening programme for breast cancer. A mammographic screening programme was attempted in 2005 but was abandoned because of lack of funding and concerns about false-positive diagnoses. Despite these barriers, national guidelines established in 2007 recommend annual mammography for women aged 40 to 49 years, and every 1 to 2 years for those aged 50 to 69 years. In a Beijing study of 1.46 million women (aged 35 to 59 years) who underwent screening by ultrasonography from 2009 to 2011, the cancer detection rate was 48.0 per 100 000, including 440 cases at early stage that constituted 69.7% of cases detected. The detection rate was lower than anticipated, maybe in part owing to the young age of the screened group and omission of mammography as a screening tool. Subsequently, a second-generation screening programme was initiated in 2012, after modification of the screening methods, cohort size (6 million), and target population that included women aged 35 to 64 years. The new screening procedures include parallel CBE and breast ultrasonography; women with suspicious findings from either examination are recommended to undergo mammographic imaging.50 Although the design of this screening protocol deviates from the standard practice of other countries, we believe that the programme will bring more research data and experience, and eventually lead to more comprehensive guidelines and consensus on a screening approach in China.

The awareness of breast cancer and acceptance of screening in Hong Kong is growing, but is still inadequate. According to the latest Breast Cancer Registry Report No 8 (2016), which covers 13 453 breast cancer patients diagnosed from 2006 onwards, the mean and median age of patients at diagnosis was 52.6 and 51.3 years, respectively, and about two-thirds of patients were aged 40 to 59 years. The screening habits among these patients were poor, with over 60% never having undergone mammography screening before their cancer diagnosis.51

Although to date there has been no population-based screening for women in Hong Kong, opportunistic screening has long been practised in the private sector. The largest voluntary self-financed and self-referred opportunistic screening programme is run by the Tung Wah Group of Hospitals. In a retrospective review of their performance from 1998 to 2002 involving over 46 600 screening mammograms, a breast cancer detection rate of five cases per 1000 population was noted, which was comparable to the detection rate of Western screening programmes at that time.52

Regarding the input of expertise and quality assurance, the Hong Kong College of Radiologists issued their mammographic statement in 2006 (latest revision in 2015).53 Quoting desirable goals recommended by the United Kingdom and United States as a reference the statement sets specific benchmarks for standards of mammographic machines, quality of screening mammograms, radiation dose limits, and accreditation requirements of reporting radiologists.53 Given these guidelines, together with recent advances in mammographic technology, we believe that there should be room for further local development of large-scale quality breast-screening programmes.

When planning a breast-screening programme, it is necessary to decide whom to screen (ie, at what age and the target screening population) and how to screen (ie, screening method).

For the decision of whom to screen, we should note that the mean age at diagnosis of breast cancer in Chinese women is 45 to 55 years, considerably younger than for western women.43 Starting screening at age 40 or 45 years would likely be a better fit for Chinese women than starting at age 50 years, as recommended by some western guidelines. As for the target screening population, current data favour universal screening over risk-based screening (pre-selecting patients according to risk profile). First, one should note that 80% of women with newly diagnosed breast cancer have no family history (ie, first-degree relative) or other significant previous risk factors, and therefore risk-based screening will miss a majority of screen-detected breast cancers.3 54 Second, a recent 10-year population-based cohort study of over 1.4 million asymptomatic Taiwanese women undergoing various breast-cancer screening regimens showed that universal mammography screening based only on age and sex was more effective than other screening regimens (risk-based biennial mammography screening or annual CBE alone).49 In that study, universal biennial mammography screening was associated with a 41% reduction in mortality and a rate of overdiagnosis of only 13%. In contrast, risk-based biennial mammography (pre-selecting patients according to risk profile or risk score) did not lead to any statistically significant reduction in mortality. Moreover, among all screening regimens, only universal biennial screening was associated with a clear downstaging shift in tumours (30% reduction of stage 2+ cancers), a crucial factor that can improve patient outcome.49

Regarding methods of screening, conventional screening uses standard two-view full-field digital (two-dimensional; 2D) mammography. Multiple studies have proven that screening by digital breast tomosynthesis (DBT; also called three-dimensional mammography) can increase cancer detection rates compared with 2D mammography alone, and can reduce the recall rate for benign findings (false-positives). 1 55 A retrospective analysis of over 454 000 screens showed that use of DBT was associated with relative increases of 41% in invasive cancer detection, 49% in positive predictive value (PPV) for recall, and 21% in PPV for biopsy, in addition to a 15% reduction in the overall number of recalls.56 A recent meta-analysis by a Korean group also showed that screening with DBT increased detection of early invasive cancers of <2 cm.57 The American College of Radiology Commission on Breast Imaging now recommends that mammography and DBT are “usually appropriate” for screening of average-risk women, noting that DBT addresses some limitations of standard digital mammography.58 In Hong Kong, DBT has been increasingly adopted to replace or serve as an adjunct to 2D mammography in opportunistic screening. We anticipate that the shift to DBT screening will become a global trend.

The use of whole-breast ultrasonography to screen dense breasts is also commonly adopted in Asia, including for opportunistic screening in Hong Kong. In Japan, this practice was reinforced by a government-funded RCT (J-START) that studied the use of adjunctive ultrasonography to supplement mammography in screening over 70 000 women. The J-START study showed favourable results of increased sensitivity and detection rate for early, preclinical cancers.41

Screening for high-risk women is often considered a separate entity. According to the American College of Radiology’s Appropriateness Criteria, women at high risk due to prior mantle radiation between the ages of 10 and 30 years should start mammography 8 years after radiation therapy, but not before age 25. For women with a genetic predisposition, annual screening mammography is recommended to begin 10 years earlier than the age that an affected relative had been diagnosed, but not before age 30. Annual screening by magnetic resonance imaging is recommended in high-risk women as an adjunct to mammography.59

We believe that health care in Hong Kong should have the capability and expertise to roll out quality, large-scale population-screening programmes that are comparable to those in other developed Asian countries and cities. When we examine the common themes among available guidelines, literature, and expert reviews worldwide, the global trend is to provide women with an informed choice.

In the discussion of whether breast-cancer screening is feasible, one should bear in mind that this is an emotive issue. Apart from the critical appraisal of scientific evidence, the interpretation of literature and subsequent formulation of recommendations should always account for the socioeconomic, historical, and contextual realities. The value judgement of women should also be respected.

Frontline experts, including breast surgeons, oncologists, breast radiologists, and their representative professional associations should all participate in guideline panels, with a will to end the ‘mammography wars’. Our Holy Grail should always be focused on improving cancer detection, reducing mortality, and improving patient outcome.

1. Eby PR. Evidence to support screening women annually. Radiol Clin North Am 2017;55:441-56.

2. Society of Breast Imaging Breast Screening Leadership Group. Screening in the 40-49 Age Group. Available from: https://www.sbi-online.org/RESOURCES/BreastScreeningLeadershipGroupResources.aspx. Accessed 19 Nov 2017.

3. Ray KM, Price ER, Joe BN. Evidence to support screening women in their 40s. Radiol Clin North Am 2017;55:429-39.

4. The American College of Radiology. Guidelines for Breast Cancer Screening—An Update. SBI Breast Imaging Symposium 2016. Available from: https://www.sbi-online.org/Portals/0/Breast%20Imaging%20Symposium%202016/Final%20Presentations/4-7%20800am%20Smith%20-%20Guidelines%20for%20Breast%20Cancer%20Screening.pdf. Accessed 19 Nov 2017.

5. Shieh Y, Eklund M, Sawaya GF, et al. Population-based screening for cancer: hope and hype. Nat Rev Clin Oncol 2016;13:550-65. CrossRef

6. Byrne SK. What's the Buzz: Tell me what’s happening in breast cancer screening. Asia Pac J Oncol Nurs 2017;4:122-6. CrossRef

7. Freer P, Moy L, DeMartini W; the Screening Leadership Group. Breast Cancer Screening: Understanding the Randomised Controlled Trials. Available from: https://www.sbi-online.org/Portals/0/Screening%20Leaders/Breast%20Cancer%20Screening_Understanding%20the%20Randomized%20Controlled%20Trials.pdf. Accessed 19 Nov 2017.

8. Duffy SW, Chen TH, Smith RA, Yen AM, Tabar L. Real and artificial controversies in breast cancer screening. Breast Cancer Manage 2013;2:519-28. CrossRef

9. Newell M, Eby PR; the Breast Screening Leadership Group. Benefits of Screening Mammography: Data from Population Service Screening. Available from: https://www.sbi-online.org/Portals/0/Screening Leaders/Benefitsof Screening Mammography_Data from PopulationService Screening.pdf. Accessed 19 Nov 2017.

10. Miller AB, Baines CJ, To T, et al. Canadian National Breast Screening Study: 2. Breast cancer detection and death rates among women aged 50 to 59 years. CMAJ 1992;147:1477-88.

11. Miller AB, Baines CJ, To T, et al. Canadian National Breast Screening Study: 1. Breast cancer detection and death rates among women aged 40 to 49 years. CMAJ 1992;147:1459-76.

12. Miller AB, To T, Baines CJ, et al. Canadian National Breast Screening Study-2: 13-year results of a randomized trial in women aged 50-59 years. J Natl Cancer Inst 2000;92:1490-9. CrossRef

13. Miller AB, To T, Baines CJ, et al. The Canadian National Breast Screening Study-1: breast cancer mortality after 11 to 16 years of follow-up. A randomized screening trial of mammography in women age 40 to 49 years. Ann Intern Med 2002;137:305-12. CrossRef

14. Miller AB, Wall C, Baines CJ, et al. Twenty five year follow-up for breast cancer incidence and mortality of the Canadian National Breast Screening Study: randomised screening trial. BMJ 2014;348:g366. CrossRef

15. Kopans DB, Feig SA. The Canadian National Breast Screening Study: a critical review. AJR Am J Roentgenol 1993;161:755-60. CrossRef

16. Burhenne LJ, Burhenne HJ. The Canadian National Breast Screening Study: a Canadian critique. AJR Am J Roentgenol 1993;161:761-3. CrossRef

17. Kopans D. Breast Imaging. 3rd ed. Lippincott Williams & Wilkins: Philadelphia; 2007.

18. Heywang-Köbrunner SH, Schreer I, Hacker A, et al. Conclusions for mammography screening after 25-year follow-up of the Canadian National Breast Cancer Screening Study (CNBSS). Eur Radiol 2016;26:342-50. CrossRef

19. Boyd NF. The review of randomization in the Canadian National Breast Screening Study. Is the debate over? CMAJ 1997;156:207-9.

20. Baines CJ, Miller AB, Kopans DB, et al. Canadian National Breast Screening Study: assessment of technical quality by external review. AJR Am J Roentgenol 1990;155:743-7; discussion 8-9. CrossRef

21. International Agency for Cancer on Research (IARC), World Health Organization. IARC Handbooks of Cancer Prevention. Volume 7: Breast Cancer Screening. IARC Press; 2002. Available from: http://www.iarc.fr/en/publications/pdfs-online/prev/handbook7/Handbook7_Breast-4.pdf. Accessed 15 Mar 2018.

22. Gotzsche PC, Olsen O. Is screening for breast cancer with mammography justifiable? Lancet 2000;355:129-34. CrossRef

23. Tabár L, Dean PB, Cen TH, et al. The impact of mammography screening on the diagnosis and management of early-phase breast cancer. In: Francescatti D, Silverstein M, editors. Breast Cancer: A New Era in Management. Springer New York; 2014: 31-78. CrossRef

24. Bock K, Borisch B, Cawson J, et al. Effect of population-based screening on breast cancer mortality. Lancet 2011;378:1775-6. CrossRef

25. Patnick J, Perry N, de Wolf C. Effect of population-based screening on breast cancer mortality—Authors’ reply. Lancet 2012;379:1298.

26. Mcneil DG. Confronting cancer: scientist at work—Peter Gotzsche; A career that bristles with against-the-grain conclusions. Available from: http://www.nytimes.com/2002/04/09/science/confronting-cancer-scientist-work-peter-gotzsche-career-that-bristles-with.html. Accessed 19 Nov 2017.

27. Cuzick J. Breast cancer screening—time to move forward. Lancet 2012;379:1289-90. CrossRef

28. Chiolero A, Rodondi N. Lessons from the Swiss Medical Board recommendation against mammography screening programs. JAMA Intern Med 2014;174:1541-2. CrossRef

29. Biller-Andorno N, Juni P. Abolishing mammography screening programs? A view from the Swiss Medical Board. N Engl J Med 2014;370:1965-7. CrossRef

30. Helvie MA, Chang JT, Hendrick RE, et al. Reduction in late-stage breast cancer incidence in the mammography era: Implications for overdiagnosis of invasive cancer. Cancer 2014;120:2649-56. CrossRef

31. Kopans DB, Smith RA, Duffy SW. Mammographic screening and “overdiagnosis”. Radiology 2011;260:616-20. CrossRef

32. Puliti D, Duffy SW, Miccinesi G, et al. Overdiagnosis in mammographic screening for breast cancer in Europe: a literature review. J Med Screen 2012;19 Suppl 1:42-56. CrossRef

33. Duffy SW, Agbaje O, Tabar L, et al. Overdiagnosis and overtreatment of breast cancer: estimates of overdiagnosis from two trials of mammographic screening for breast cancer. Breast Cancer Res 2005;7:258-65. CrossRef

34. Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med 2012;367:1998-2005. CrossRef

35. Duffy SW, Dibden A, Michalopoulos D, et al. Screen detection of ductal carcinoma in situ and subsequent incidence of invasive interval breast cancers: a retrospective population-based study. Lancet Oncol 2016;17:109-14. CrossRef

36. Tosteson AN, Fryback DG, Hammond CS, et al. Consequences of false-positive screening mammograms. JAMA Intern Med 2014;174:954-61. CrossRef

37. Schwartz LM, Woloshin S, Fowler FJ, Jr, et al. Enthusiasm for cancer screening in the United States. JAMA 2004;291:71-8. CrossRef

38. Kopans DB. An open letter to panels that are deciding guidelines for breast cancer screening. Breast Cancer Res Treat 2015;151:19-25. CrossRef

39. Duffy SW, Chen TH, Smith RA, Yen AM, Tabar L. Real and artificial controversies in breast cancer screening. Breast Cancer Manage 2013;2:519-28. CrossRef

40. Montero AJ, Eapen S, Gorin B, et al. The economic burden of metastatic breast cancer: a U.S. managed care perspective. Breast Cancer Res Treat 2012;134:815-22. CrossRef

41. Ohuchi N, Ishida T, Kawai M, et al. Randomized controlled trial on effectiveness of ultrasonography screening for breast cancer in women aged 40-49 (J-START): research design. Jpn J Clin Oncol 2011;41:275-7. CrossRef

42. Sung H, Rosenberg PS, Chen WQ, et al. Female breast cancer incidence among Asian and Western populations: more similar than expected. J Natl Cancer Inst 2015;107. CrossRef

43. Fan L, Strasser-Weippl K, Li JJ, et al. Breast cancer in China. Lancet Oncol 2014;15:e279-89. CrossRef

44. Youlden DR, Cramb SM, Yip CH, et al. Incidence and mortality of female breast cancer in the Asia-Pacific region. Cancer Biol Med 2014;11:101-15.

45. Singapore Health Promotion Board. National Registry of Diseases Office. Singapore Cancer Registry Interim Annual Report: Trends in cancer incidence in Singapore 2010-2014. Available from: https://www.nrdo.gov.sg/docs/librariesprovider3/default-document-library/cancer-trends-2010-2014_interim-annual-report_final-%28public%29.pdf?sfvrsn=0. Accessed 19 Nov 2017.

46. Hospital Authority, Hong Kong. Hong Kong Cancer Registry. Available from: http://www3.ha.org.hk/cancereg/. Accessed 19 Mar 2018.

47. Leung GM, Thach TQ, Lam TH, et al. Trends in breast cancer incidence in Hong Kong between 1973 and 1999: an age-period-cohort analysis. Br J Cancer 2002;87:982-8. CrossRef

48. Breast cancer in Asia—The Challenge and Response. A Report from the Economist Intelligence Unit. Available from: https://www.eiuperspectives.economist.com/sites/default/files/EIU Breast Cancer in Asia_Final.pdf. Accessed 19 Nov 2017.

49. Yen AM, Tsau HS, Fann JC, et al. Population-based breast cancer screening with risk-based and universal mammography screening compared with clinical breast examination: a propensity score analysis of 1 429 890 Taiwanese women. JAMA Oncol 2016;2:915-21. CrossRef

50. Song QK, Wang XL, Zhou XN, et al. Breast cancer challenges and screening in China: lessons from current registry data and population screening studies. Oncologist 2015;20:773-9. CrossRef

51. A Hong Kong Breast Cancer Foundation Initiative. Hong Kong Breast Cancer Registry Report No. 8, 2016. Available from: http://www.hkbcf.org/download/bcr_report8/hkbcf_report_2016_full_report.pdf. Accessed 19 Nov 2017.

52. Lui CY, Lam HS, Chan LK, et al. Opportunistic breast cancer screening in Hong Kong; a revisit of the Kwong Wah Hospital experience. Hong Kong Med J 2007;13:106-13.

53. Hong Kong College of Radiologists Mammography Statement. Available from: https://www.hkcr.org/templates/OS03C00336/case/lop/HKCR MammographyStatement_rev20150825.pdf. Accessed 19 Nov 2017.

54. Destounis SV, Arieno AL, Morgan RC, et al. Comparison of breast cancers diagnosed in screening patients in their 40s with and without family history of breast cancer in a community outpatient facility. AJR Am J Roentgenol 2014;202:928-32. CrossRef

55. Rafferty EA, Rose SL, Miller DP, et al. Effect of age on breast cancer screening using tomosynthesis in combination with digital mammography. Breast Cancer Res Treat 2017;164:659-66. CrossRef

56. Friedewald SM, Rafferty EA, Rose SL, et al. Breast cancer screening using tomosynthesis in combination with digital mammography. JAMA 2014;311:2499-507. CrossRef

57. Yun SJ, Ryu CW, Rhee SJ, et al. Benefit of adding digital breast tomosynthesis to digital mammography for breast cancer screening focused on cancer characteristics: a meta-analysis. Breast Cancer Res Treat 2017;164:557-69. CrossRef

58. Monticciolo DL, Newell MS, Hendrick RE, et al. Breast cancer screening for average-risk women: recommendations from the ACR commission on breast imaging. J Am Coll Radiol 2017;14:1137-43. CrossRef

59. Expert Panel on Breast Imaging; Mainiero MB, Moy L, Baron P, et al. ACR Appropriateness Criteria Breast Cancer Screening. J Am Coll Radiol 2017;14:S383-90. CrossRef

Ketamine is an N-methyl-d-aspartate receptor antagonist, a dissociative anaesthetic agent and a treatment option for major depression, treatment-resistant depression, and bipolar disorder. Its strong psychostimulant properties and easy absorption make it a favourable candidate for substance abuse. Ketamine entered Hong Kong as a club drug in 2000 and the first local report of ketamine-associated urinary cystitis was published in 2007. Ketamine-associated lower–urinary tract symptoms include frequency, urgency, nocturia, dysuria, urge incontinence, and occasionally painful haematuria. The exact prevalence of ketamine-associated urinary cystitis is difficult to assess because the abuse itself and many of the associated symptoms often go unnoticed until a very late stage. Additionally, upper–urinary tract pathology, such as hydronephrosis, and other complications involving neuropsychiatric, hepatobiliary, and gastrointestinal systems have also been reported. Gradual improvement can be expected after abstinence from ketamine use. Sustained abstinence is the key to recovery, as relapse usually leads to recurrence of symptoms. Both medical and surgical management can be used. The Youth Urological Treatment Centre at the Prince of Wales Hospital, Hong Kong, has developed a four-tier treatment protocol with initial non-invasive investigation and management for these patients. Multidisciplinary care is essential given the complex and diverse psychological factors and sociological background that underlie ketamine abuse and abstinence status.

Ketamine is an N-methyl-d-aspartate (NMDA) receptor antagonist, a dissociative anaesthetic agent that was first synthesised in the United States in 1962. It has been widely used in both human and veterinary medicine since 1971. It has also been used as a treatment for major depression, treatment-resistant depression, and bipolar disorder.1 However, its strong psychostimulant properties and easy absorption make it a favourable candidate for substance abuse. Ketamine abuse has become increasingly common over the past two decades. It entered Hong Kong as a club drug in 2000 and was initially used as a ‘top-up’ drug to ecstasy (3,4-methylenedioxymethamphetamine) by those aiming to elevate and redirect the ‘high’.2 It was viewed as a poor man’s version of cocaine, as it is available in powder form and can be consumed by snorting. By 2002, ketamine had become the drug of choice instead of a ‘top-up’ drug. Users self-administered ketamine to have a ‘time-out’ or ‘sit-down-to-float’ experience.2 3 Within a short period, the number of reported ketamine abusers in Hong Kong increased from 1605 in 2000 to its peak of 5280 in 2009. Ketamine remained the most popular psychotropic substance abused from 2005 to 2014 (Fig 1).4

The abuse of ketamine and its popularity nonetheless created a new medical entity. The first report of ketamine-associated urinary cystitis in Hong Kong was published in 2007.5 In the same year, Shahani et al reported a similar condition overseas.6 In the past decade, owing to the joint efforts of urologists, general surgeons, physicians, psychiatrists, pathologists, and basic scientists, we have gained a better understanding of other ketamine-associated conditions. This understanding spans from pathology to clinical management, from urological complications to upper-gastrointestinal (GI) complications, and from a mouse model to humans. We have also explored holistic ways to manage this condition in the long term, such as helping young adults to have a fresh start while living with potentially irreversible complications. This article reviews the evolution of local clinical awareness and management of these complications, with a particular focus on the work and contributions of local researchers.

Chu et al5 reported the first local case series of ketamine-associated bladder dysfunction in 2007. Ketamine-associated lower–urinary tract symptoms (LUTS) include frequency, urgency, nocturia, dysuria, urge incontinence, and occasionally painful haematuria. The exact prevalence of ketamine-associated urinary cystitis is difficult to assess because ketamine abuse and many of the associated symptoms often go unnoticed until a very late stage. A survey among 12 000 local secondary school students revealed that 18.5% of non-psychotropic substance users had LUTS, whereas 47.8% and 60.7% of psychotropic substance users and ketamine users had LUTS, respectively.7 Unpublished data from the same study showed that compared with non-psychotropic substance users, sole ketamine users were five times as likely to have LUTS, whereas concomitant users of ketamine and methamphetamine were eight times as likely to have LUTS.

In 2015, Yee et al8 reported the largest available cohort of both active and past ketamine users who had ketamine-associated uropathy. Among 463 patients, ketamine users had a significantly higher pelvic pain and urgency/frequency (PUF) score than ex-ketamine users. The PUF score is initially used to assess interstitial cystitis, and a higher PUF score correlates with worse symptoms. Among active ketamine users, a higher PUF score was found to correlate with a poorer quality of life and a smaller functional bladder capacity.8 Achieving abstinence from ketamine use and consuming smaller amounts of ketamine were factors that predicted improvement in PUF score.

As well as bladder involvement, upper–urinary tract pathology presenting as hydronephrosis and flank pain has also been reported (Fig 2). A study by Yee et al9 of 572 patients with ketamine-related LUTS found that up to 16.8% (n=96) of patients had hydronephrosis according to ultrasonography. Hydronephrosis was frequently accompanied by ureteral lesions, ureteral wall thickening, or vesicoureteral (vesicoureteric) reflux. Similarly, Chu et al10 reported that 51% (30/59) of their patients had hydronephrosis according to ultrasonography.

Although the exact mechanism of ketamine-associated cystitis remains to be explored, there is evidence that ketamine metabolites in the urine induce chemical irritation of the urothelium, thereby causing an inflammatory response.11 Severe irritation may lead to denudation of the urothelium and consequent transmural inflammation, loss of muscle thickness, fibrosis of the detrusor muscle, and ultimately poor urinary bladder compliance. Vesicoureteral reflux or urinary stasis in the ureter may occur, causing chronic ureteral inflammation and ureteral stricture. There is also evidence that both systemic and local inflammatory markers are elevated in ketamine users.10 12 In addition, the NMDA antagonist properties of ketamine may exert their effect via a central pathway.13

As ketamine is a psychostimulant, it is not surprising that it is associated with long-term neurocognitive problems. Chan et al14 found that when ketamine users were compared with healthy controls, they had impaired verbal fluency, cognitive processing speed, and verbal learning. Heavy ketamine use correlated with deficits in verbal memory and visual recognition memory. Liang et al15 also identified predominant verbal and visual memory impairment in ketamine poly-drug users. Unfortunately, these deficits persisted in ex-users. A much higher incidence of psychiatric co-morbidities, including psychosis, depression, and anxiety, was observed among ketamine users.16

Structural brain damage associated with ketamine abuse was supported by magnetic resonance imaging (MRI) by Wang et al.17 They were the first group to report patches of degeneration in the superficial white matter as early as 1 year after ketamine addiction onset. Cortical atrophy was also found in the frontal, parietal, or occipital cortices of addicts.17 Another MRI study also provided evidence of brain damage in chronic ketamine users. Reduced grey- and white-matter volumes were noted in the bilateral orbitofrontal cortex, right medial prefrontal cortex, and bilateral hippocampi. There was also significantly decreased connectivity inside the brain in chronic ketamine abusers.18

A series of studies on the neurotoxicity of ketamine suggested that ketamine could cause apoptosis of neuronal cells in both in-vitro and in-vivo models. Ketamine also potentially causes phosphorylation of tau protein, a marker of Alzheimer’s degeneration in the brain.19 20 21 22

In 2009, Wong et al23 reported ketamine-associated hepatobiliary complications for the first time. Three ketamine abusers presented with recurrent epigastric pain and dilated bile ducts mimicking choledochal cysts. Subsequently, more similar cases were identified. Fusiform dilatation of the common bile duct was also observed.24 25 26 Liver biopsy confirmed development of active liver and/or bile duct injury. A study of 297 chronic ketamine abusers with urinary tract dysfunction showed that the prevalence of liver injury was 9.8%.27 These studies and reports show the possibility and severity of damage by ketamine to the hepatobiliary and pancreatic system.

The exact mechanism of ketamine-associated bile injury is still unknown. The associated rise in C-reactive protein suggests a possible inflammatory process in the liver parenchyma, including or excluding the bile duct.27 Others have postulated that either central or direct action of ketamine on the biliary smooth muscle in turn leads to the cholestasis and biliary dilatation observed in ketamine abusers.28

In addition to urological complaints, GI problems are also frequently the symptoms for which ketamine abusers seek medical help. A review of 233 ketamine-related visits to accident and emergency departments found that 49 (21.0%) patients had abdominal pain, 23 (9.9%) had nausea or vomiting, and 41 (17.6%) had abdominal tenderness.29 Gastrointestinal complaints often co-exist with and precede the presentation of urological symptoms. Liu et al30 found that about a quarter (168; 27.5%) of 611 ketamine users who sought treatment for ketamine uropathy reported the presence of upper-GI symptoms, whereas only 42 (5.2%) of 804 non-ketamine users attending a general urology clinic reported similar symptoms (P<0.001). The majority of the symptoms reported were epigastric pain and recurrent vomiting. Nearly three-quarters of patients required hospitalisation for acute or chronic upper-GI symptoms. With the exception of acid reflux and perforated peptic ulcer, the prevalence of all the above-mentioned symptoms and hospitalisation rates were statistically significantly higher in ketamine users than in non-ketamine users. All 168 patients using ketamine had undergone oesophagogastroduodenoscopy during which biopsies were taken. Pathological findings ranged from gastritis to gastroduodenal erosions, peptic ulceration, and intestinal metaplasia.30

Liu et al30 also found that more than 80% of patients developed upper-GI symptoms before urological symptoms. Patients developed upper-GI symptoms after a mean (standard deviation) of 5.1 (3.1) years of ketamine use and developed uropathy symptoms after another 4.4 (3.0) years of ketamine use.30 Epigastric symptoms are not common in young people, but common in ketamine abusers. This difference may provide an opportunity to identify hidden ketamine abuse when assessing young patients with epigastric symptoms. The identification of ketamine use is important, as cessation of use can greatly improve GI symptoms.31 Further referral for help and counselling may improve psychological and physical health and promote long-term ketamine abstinence.

The exact pathophysiological mechanism by which ketamine produces upper-GI toxicity remains unknown but there are several postulations. First, ketamine, as an NMDA antagonist, might act on local smooth muscle or the central nervous system, thereby affecting gastric motility and leading to cramping pain. Second, microvascular damage by ketamine and its metabolites, which was believed to be a possible cause of ketamine uropathy, might also cause similar microvascular damage in the stomach and duodenum, leading to ischaemic pain and inflammation. Likewise, circulating ketamine might also trigger some unknown autoimmune responses, and thus induce interstitial inflammation in the urinary and GI tracts. Finally, as many ketamine abusers like to swallow the nasal drips occurring from ketamine inhalation, the swallowed ketamine might also induce direct cytotoxic injury to the vulnerable GI tract.30

As in the management of other substance abuse, abstinence is the key to success in overall management of ketamine use. Whereas other treatment modalities may relieve symptoms and hasten the recovery process, many ketamine abusers have complicated underlying psychosocial problems and psychiatric co-morbidities. Long-term and consistent support and encouragement from doctors, nurses, social workers, family, and friends are vital for success.

Recurrence of symptoms after resuming ketamine use highlights the importance of ketamine abstinence. Studies have shown that abstinence leads to symptomatic improvement. Compared with active ketamine abusers, those who had abstained for 1 year had significantly lower PUF scores and a larger voided volume. There was a trend towards higher voided volumes and lower PUF scores as duration of ketamine cessation increased, although neither variable was statistically significant.32 Another follow-up study of 101 participants who had abstained from ketamine and 218 active ketamine users showed that the abstinence group had a statistically significantly lower PUF score, and a higher functional bladder capacity.8 Moreover, abstinence was the only protective factor associated with fewer symptoms, larger voided volume, and bladder capacity.33

Nonetheless, abstinence does not lead to immediate and full recovery of symptoms. Gradual improvement can be expected but sustained abstinence is the key to recovery. Patience and continuous support are of paramount importance. A study showed that on admission to a drug rehabilitation centre, 90% of 40 female ex-ketamine users still had active urinary symptoms, with increased 24-hour urinary frequency, lower maximum voided volume, smaller median functional bladder capacity, and higher mean Urogenital Distress Inventory Short Form (UDI-6) and Incontinence Impact Questionnaire Short Form (IIQ-7) scores, when compared with age-matched controls who attended a general gynaecology clinic. After having stopped using ketamine for 3 months or more, mean 24-hour urinary frequency and mean UDI-6 and IIQ-7 scores decreased, and maximum voided volume increased. These scores further improved after another 3 months, although this group continued to perform more poorly in all aspects compared with controls.34

As ketamine-associated uropathy is an evolving ‘disease entity’, the exact pathophysiology remains to be elucidated. Some of the clinical features share similarities with interstitial cystitis. Protocols are being developed to cater to the needs of patients in Hong Kong. A one-stop service model has been adopted by the Youth Urological Treatment Centre at the Prince of Wales Hospital since 2011 (Fig 3). The standard treatment protocol involves four tiers of treatment, starting with an initial non-invasive investigative approach, including questionnaire assessment of symptoms and calculation of (1) functional bladder capacity by measuring voided volume using uroflowmetry and (2) residual urine using ultrasound bladder scanning. This non-invasive investigative approach helps gain patients’ trust and improve adherence to later follow-up.

First-tier treatment includes oral non-steroidal anti-inflammatory drugs (NSAIDs) (eg, diclofenac) and anticholinergics (eg, solifenacin) or COX-II inhibitors (eg, etoricoxib) if patients cannot tolerate NSAIDS. Simple analgesics such as paracetamol and phenazopyridine are used for pain control. The Youth Urological Treatment Centre has reported the largest series of patients with ketamine-associated uropathy and their corresponding outcomes. Of 290 patients with ketamine cystitis who received first-line treatment, 202 (69.7%) reported symptom improvement and a reduction in PUF scores. Functional bladder capacity was also shown to have improved.8

The opioid group of analgesics and pregabalin are used in the next tier of pain-control treatment when first-tier treatment is insufficient for symptom relief. Sixty-two patients received second-line treatment and 42 (67.7%) responded to treatment.8

Third-tier treatment consists of a course of intravesical instillation of sodium hyaluronate (6-weekly instillations followed by 2-monthly instillations) attempting to repair the glycosaminoglycan layer. The drug is given to patients whose symptoms remain uncontrolled after second-tier treatment. Seventeen patients in the cohort received the third-tier treatment and eight completed the course. Significant improvement in voided volume was noted and five were able to reduce their oral medication usage after treatment. No significant adverse effects were reported.8

Unfortunately, for a proportion of patients with extremely refractory symptoms, surgery becomes the fourth-tier treatment of choice. In the Youth Urological Treatment Centre series, one patient in the cohort required hydrodistension and another underwent robotic-assisted laparoscopic augmentation cystoplasty. The patient with hydrodistension experienced a recurrence of symptoms post-treatment.8 Ng et al35 reported on four patients who underwent augmentation cystoplasty. Although they showed initial improvement, all patients relapsed and resumed ketamine use postoperatively. Three of the patients showed a further deterioration in renal function, secondary to new-onset ureteral strictures and/or sepsis. Therefore, patient selection, education, close follow-up, and support are vital to the success of augmentation cystoplasty.35

Given the complex and diverse psychological factors and sociological background contributing to an individual’s decision to abuse ketamine or achieve abstinence, joint multidisciplinary efforts are required to help affected young adults. Doctors, social workers, teachers, psychiatrists, psychologists, nurses, and patients’ families all need to support them on their long road to recovery, to help them rehabilitate physically and achieve sustained abstinence from ketamine.33 36 37

Since the initial discovery of ketamine-associated uropathy, the impact of this disease entity has become more prominent in Asian countries. Thanks to the joint efforts of urologists, gynaecologists, surgeons, psychiatrists, pathologists, and social workers, as well as the support of local government, the extent of medical complications has been revealed to also involve the brain, liver, and GI system. Many ketamine abusers are ‘hidden’ and can use ketamine stealthily at home for years without their family noticing. Clinicians must take the opportunity to identify hidden abusers when they consult for non-specific symptoms such as epigastric pain and LUTS. Doing so will not only enable early diagnosis of ketamine-associated uropathy, but it will also help provide appropriate medical treatment in a timely manner. In addition to medical therapy, referral for appropriate psychosocial support is essential to sustain abstinence and manage underlying psychosocial problems.

The Youth Urological Treatment Centre was developed by joint efforts of The Chinese University of Hong Kong and the Hong Kong Hospital Authority, with generous support from the Beat Drugs Fund of the Narcotics Division, Security Bureau, Government of the Hong Kong Special Administrative Region.

The authors have no conflicts of interest to disclose.

1. Hasselmann HW. Ketamine as antidepressant? Current state and future perspectives. Curr Neuropharmacol 2014;12:57-70. CrossRef

2. Joe Laidler KA. The rise of club drugs in a heroin society: the case of Hong Kong. Subst Use Misuse 2005;40:1257-78. CrossRef

3. Joe-Laidler K, Hunt G. Sit down to float: the cultural meaning of ketamine use in Hong Kong. Addict Res Theory 2008;16:259-71. CrossRef

4. Narcotics Division, Security Bureau, HKSAR Government. Statistical tables customization. Available from: http://cs-crda.nd.gov.hk/en/enquiry.php. Accessed 22 Sep 2017.

5. Chu PS, Kwok SC, Lam KM, et al. ‘Street ketamine’–associated bladder dysfunction: a report of ten cases. Hong Kong Med J 2007;13:311-3.

6. Shahani R, Streutker C, Dickson B, et al. Ketamine-associated ulcerative cystitis: a new clinical entity. Urology 2007;69:810-2. CrossRef

7. Tam YH, Ng CF, Wong YS, et al. Population-based survey of the prevalence of lower urinary tract symptoms in adolescents with and without psychotropic substance abuse. Hong Kong Med J 2016;22:454-63. CrossRef

8. Yee CH, Lai PT, Lee WM, et al. Clinical outcome of a prospective case series of patients with ketamine cystitis who underwent standardized treatment protocol. Urology 2015;86:236-43. CrossRef

9. Yee CH, Teoh YC, Lai PT, et al. The risk of upper urinary tract involvement in patients with ketamine-associated uropathy. Int Neurourol J 2017;21:128-32. CrossRef

10. Chu PS, Ma WK, Wong SC, et al. The destruction of the lower urinary tract by ketamine abuse: a new syndrome? BJU Int 2008;102:1616-22. CrossRef

11. Chu PS, Ng CF, Ma WK. Ketamine uropathy: Hong Kong experience. In: Yew DT, editor. Ketamine: Use and Abuse. CRC Press; 2015: 207-26.

12. Division of Urology, Department of Surgery, Princess Margaret Hospital and Tuen Mun Hospital. Research on Urological Sequelae of Ketamine Abuse. Available from: http://www.nd.gov.hk/pdf/20110307_beat_drug_fund_report.pdf. Accessed 22 Sep 2017.

13. Jankovic SM, Jankovic SV, Stojadinovic D, et al. Effect of exogenous glutamate and N-methyl-D-aspartic acid on spontaneous activity of isolated human ureter. Int J Urol 2007;14:833-7. CrossRef

14. Chan KW, Lee TM, Siu AM, et al. Effects of chronic ketamine use on frontal and medial temporal cognition. Addict Behav 2013;38:2128-32. CrossRef

15. Liang HJ, Lau CG, Tang A, et al. Cognitive impairments in poly-drug ketamine users. Addict Behav 2013;38:2661-6. CrossRef

16. Tang WK, Morgan CJ, Lau GC, et al. Psychiatric morbidity in ketamine users attending counselling and youth outreach services. Subst Abus 2015;36:67-74. CrossRef

17. Wang C, Zheng D, Xu J, et al. Brain damages in ketamine addicts as revealed by magnetic resonance imaging. Front Neuroanat 2013;7:23. CrossRef

18. Narcotics Division, Security Bureau, HKSAR Government. Evidence of Brain Damage in Chronic Ketamine Users—a Brain Imaging Study. Available from: http://www.nd.gov.hk/pdf/Evidence-of-Brain-Damage-in-Chronic-Ketamine-Users_Final-report.pdf. Accessed 22 Sep 2017.

19. Yeung LY, Wai MS, Fan M, et al. Hyperphosphorylated tau in the brains of mice and monkeys with long-term administration of ketamine. Toxicol Lett 2010;193:189-93. CrossRef

20. Narcotics Division, Security Bureau, HKSAR Government. Long-term Ketamine Abuse and Apoptosis in Cynomolgus Monkeys and Mice. Available from: http://www.nd.gov.hk/pdf/long_term_ketamine_abuse_apoptosis_in_cynomologus_monkeys_and_mice/final_report_with_all_attachments.pdf. Accessed 22 Sep 2017.

21. Tan S, Rudd JA, Yew DT. Gene expression changes in GABA(A) receptors and cognition following chronic ketamine administration in mice. PLoS ONE 2011;6:e21328. CrossRef

22. Mak YT, Lam WP, Lü L, et al. The toxic effect of ketamine on SH-SY5Y neuroblastoma cell line and human neuron. Microsc Res Tech 2010;73:195-201.

23. Wong SW, Lee KF, Wong J, et al. Dilated common bile ducts mimicking choledochal cysts in ketamine abusers. Hong Kong Med J 2009;15:53-6.

24. Ng SH, Lee HK, Chan YC, et al. Dilated common bile ducts in ketamine abusers. Hong Kong Med J 2009;15:157.

25. Cheung TT, Poon RT, Chan AC, et al. Education and imaging. Hepatobiliary and pancreatic: cholangiopathy in ketamine user—an emerging new condition. J Gastroenterol Hepatol 2014;29:1663. CrossRef

26. Lui KL, Lee WK, Li MK. Ketamine-induced cholangiopathy. Hong Kong Med J 2014;20:78.e1-2. CrossRef

27. Wong GL, Tam YH, Ng CF, et al. Liver injury is common among chronic abusers of ketamine. Clin Gastroenterol Hepatol 2014;12:1759-62.e1. CrossRef

28. Lo RS, Krishnamoorthy R, Freeman JG, et al. Cholestasis and biliary dilatation associated with chronic ketamine abuse: a case series. Singapore Med J 2011;52:e52-5.

29. Ng SH, Tse ML, Ng HW, et al. Emergency department presentation of ketamine abusers in Hong Kong: a review of 233 cases. Hong Kong Med J 2010;16:6-11.

30. Liu SY, Ng SK, Tam YH, et al. Clinical pattern and prevalence of upper gastrointestinal toxicity in patients abusing ketamine. J Dig Dis 2017;18:504-10. CrossRef

31. Poon TL, Wong KF, Chan MY, et al. Upper gastrointestinal problems in inhalational ketamine abusers. J Dig Dis 2010;11:106-10. CrossRef

32. Mak SK, Chan MT, Bower WF, et al. Lower urinary tract changes in young adults using ketamine. J Urol 2011;186:610-4. CrossRef

33. Tam YH, Ng CF, Pang KK, et al. One-stop clinic for ketamine-associated uropathy: report on service delivery model, patients’ characteristics and non-invasive investigations at baseline by a cross-sectional study in a prospective cohort of 318 teenagers and young adults. BJU Int 2014;114:754-60. CrossRef

34. Cheung RY, Chan SS, Lee JH, et al. Urinary symptoms and impaired quality of life in female ketamine users: persistence after cessation of use. Hong Kong Med J 2011;17:267-73.

35. Ng CF, Chiu PK, Li ML, et al. Clinical outcomes of augmentation cystoplasty in patients suffering from ketamine-related bladder contractures. Int Urol Nephrol 2013;45:1245-51. CrossRef

36. Ng CF. Editorial comment to possible pathophysiology of ketamine-related cystitis and associated treatment strategies. Int J Urol 2015;22:826. CrossRef

37. Chu PS, Dasgupta P. Lessons learned from Asian Urology. BJU Int 2014;114:633. CrossRef