 |
 |
|
|
|
 Late-life depression is a major public health problem. Major depression occurs in 1% to 20% of the general elderly population (1–4), while clinically significant depression occurs in up to 20% of seniors living in the community (2) and 37% of those treated in primary care (5). Depressive disorders in the elderly are associated with increased risks for impaired physical and emotional health (2), pain (6), loss of productivity (2), mortality from medical conditions (7) and suicide (2), and increased use of medical and mental health services (2,8–10). Late-life depression is associated with poor prognosis: one metaanalysis of elderly individuals suffering from depression in the community and in primary care reported that, at 24-month follow-up, only 33% were well, 33% suffered from depression, and 21% had died (5). Given the pervasiveness and impact of late-life depression, delivery of effective treatment is essential to improving quality and length of life. Most seniors with depression are treated in primary care (11). Many receive inadequate treatment (3,12–17), although clinical practice guidelines have been developed to assist practitioners with treatment. In 2001, the CPA released its Clinical Guidelines for the Treatment of Depressive Disorders (18) that included recommendations tailored to individuals aged 65 years and older. Effective treatments included pharmacotherapy, electroconvulsive therapy for patients with significant medical comorbidity and less tolerability to pharmacotherapy, and psychotherapy for mild-to-moderate depression (18, p 63S–76S). The Guidelines also specified first-, second- and third-line pharmacotherapies, all of which are antidepressants. As the Canadian population ages, the issue of receiving appropriate treatment for late-life depression assumes greater importance. We undertook this research to compare the pharmacotherapies dispensed in response to a new depression diagnosis with existing guidelines. Our study aimed to determine the factors predicting receipt of psychoactive medication and first-line pharmacotherapy in Quebec seniors with new episodes of depression. Identifying the determinants of appropriate treatment is the first step to reducing disability, poor quality of life, increased risk for death, and avoidable use of medical and mental health care services. MethodsStudy Design and Data Source We conducted a population-based, retrospective cohort study based on data we obtained from the databases of the RAMQ (the Quebec health insurance board), which remunerates physicians for services and administers the Quebec public drug insurance plan. In 2002, the public drug insurance plan covered approximately 871 800 residents aged 65 to 85 years (19). Using unique encrypted numbers for patients and physicians to maintain anonymity, we linked 3 RAMQ databases at the patient level. We obtained data on patient age, sex, and region of residence from the beneficiaries database. Diagnoses and diagnosing physician specialty were obtained from the medical services database. Drug name, type of medication, quantity and concentration, date prescription filled, and prescribing physician specialty were obtained from the pharmaceutical database. The latter database is based on prescription claims for drugs dispensed and has been shown to be reliable and valid (20). Identification of Study Subjects We identified beneficiaries aged 65 to 84 years and having a new diagnosis of depressive disorder given by either a primary care physician (for example a family medicine physician, internist, or geriatrician) or a psychiatrist between October 1, 2000, and March 31, 2001 (a 6-month case identification period). We obtained all pharmaceutical records from October 1, 1999, to March 31, 2002. There were 2 observation periods: first, a follow-up 1-year period from the date of the first diagnostic coding of depression, during which subjects had to remain alive; and second, an antecedent 1-year period from the date of the first diagnostic coding of depression, during which subjects having a depression diagnosis were excluded, to retain patients with new-onset disorders. Diagnostic Codes for Depression We used the first diagnostic coding of depressive disorder during the case identification period to identify patients with depression. One diagnosis per physician visit was provided, irrespective of whether medication was prescribed. Although the RAMQ uses diagnostic codes from the ICD-9 (21), some physicians may have substituted DSM-IV (22) diagnostic codes that are more detailed for mental health disorders. To compensate for this discrepancy, we created an algorithm to retain patients with probable depression. Our physician experts identified 19 diagnostic codes that might have been used to indicate depression (Table 1); of these, we selected the codes for which diagnoses were common to both the ICD-9 and DSM-IV (for example dysthymia [300.4] and depression NOS [311.0, 311.9]).
Study MedicationsPsychoactive medications were identified from a comprehensive medication list used by the RAMQ. We assigned 9 medication classes: antidepressants, anticonvulsants, anxiolytics, barbiturates, benzodiazepines, lithium, major tranquilizers, stimulants, and others (for example, levodopa and sumatriptan). Medication use was defined as having at least one prescription claim for a study medication. New users of antidepressants were those who had no antidepressant pharmacy claim in the 365 days preceding diagnosis. Low-dose trazodone (for example daily dosage under 150 mg) was not classified as an antidepressant because it is used to aid sleep. We created a comorbidity score based on pharmacy claims data (23). Outcomes We defined receipt of psychoactive medication as having at least one dispensing claim for a study drug in the year following diagnosis, beginning with the day of diagnosis. We defined receipt of recommended first-line pharmacotherapy as having a dispensing claim for a first-line medication at the first prescribing visit, defined as the visit during which psychoactive medication was initially prescribed (which could have occurred concurrently with, or following, diagnosis). First-line pharmacotherapies listed by the CPA Guidelines (18) included the antidepressants moclobemide, citalopram, bupropion, fluvoxamine, mirtazapine, nefazodone, paroxetine, sertraline, and venlafaxine. In contrast to other late-life depression guidelines (24), fluoxetine was not considered first-line because it has a long half-life that increases the risk for drug interactions (18,25). Prior to beginning this study, we obtained approval from the University of Montreal, Faculty of Medicine Research Ethics Committee. Statistical Analysis Descriptive statistics were used to characterize the study sample. To compare subjects by sex, we used t tests and chi-square tests as appropriate. Psychoactive medication dispensed was compared by diagnosis with chi-square and Fisher’s exact tests. To account for possible dependence among different patients of the same physician, we employed a generalized estimating equations approach (26) for the 2 binary outcomes. This statistical method accounts for clustered data and for the unbalanced structure of the data, in that the number of patients varied across physicians. We used the SAS procedure GENMOD (SAS for Windows 8.02; see 27,28), which indicated that there were 2433 distinct physician clusters with patient numbers ranging from 1 to 14. Predictors of receiving psychoactive medication were evaluated among all patients in the study (n = 5258). Predictors of receiving recommended first-line pharmacotherapy were evaluated only among patients dispensed psychoactive medication (because it is possible that either psychotherapy or electroconvulsive therapy was the treatment of choice) and who had not received antidepressants in the year prior to diagnosis (n = 2623) (because it is possible that the originally dispensed antidepressant was switched if patients experienced side effects or failed to respond). For receipt of psychoactive medication, the candidate variables included patient age, sex, region of residence (urban or rural), diagnosis (dysthymia or depression NOS), and diagnosing physician specialty (primary care physician or psychiatrist). For receipt of recommended first-line pharmacotherapy, the candidate variables included these as well as having the same physician diagnosing and treating the patient (yes or no), and receiving the medication within 31 days of diagnosis (yes or no). Two-tailed P values less than 0.05 were considered statistically significant. ResultsPatient Characteristics We identified 5258 seniors with a new diagnosis of depressive disorder during the case-identification period (Table 2). Patient characteristics are presented in Table 2 according to sex. Most patients were dispensed psychoactive medication in the year following diagnosis (84.1%) and did not receive antidepressants in the preceding 365 days (64.3%). Of those given psychoactive medication, 3244 (73.4%) were diagnosed and treated by the same physician; 3840 (86.9%) received medication in the month following diagnosis; and 2623 (59.3%) had not received antidepressants in the previous year. More patients who were diagnosed and treated by the same physician received medication within the month of diagnosis, compared with those having different diagnosing and prescribing physicians (78.1% and 42.0%, respectively; P = 0.0001).
Pharmacotherapy Preceding and Following Depression Diagnosis Psychoactive medication use increased following diagnosis. In the year prior to diagnosis, patients were prescribed antidepressants (35.7%), benzodiazepines (58.2%), major tranquilizers (7.4%), anxiolytics (4.5%), anticonvulsants (2.9%), lithium (1.4%), stimulants (0.3%), and barbiturates (0.5%). In the year following diagnosis, patients were prescribed antidepressants (64.2%), benzodiazepines (61.7%), major tranquilizers (12.1%), anxiolytics (5.0%), anticonvulsants (4.6%), lithium (1.6%), stimulants (0.5%), and barbiturates (0.6%). Women, compared with men, received more antidepressants (39.1% and 28.1%, respectively; P = 0.0001), benzodiazepines (61.9% and 49.7%, respectively; P = 0.0001), and anxiolytics (5.0% and 3.4%, respectively; P = 0.0114) prior to diagnosis. Women also received more antidepressants; (63.3% and 55.6%, respectively; P = 0.0001) (66.5% and 58.8%, respectively; P = 0.0001) and benzodiazepines (64.4% and 55.6%, respectively) P = 0.0001) following diagnosis. Psychoactive Medication Dispensed at the First Prescribing Visit Antidepressants (63.6%) and benzodiazepines (45.3%) were the most commonly dispensed medications at the first prescribing visit. Patients with depression NOS received more antidepressants than did patients with dysthymia (59.9% and 42.8%, respectively; P = 0.0001). Patients with dysthymia received more benzodiazepines (50.3% and 43.7%, respectively; P = 0.0001), major tranquilizers (13.0% and 6.2%, respectively; P = 0.0001), and stimulants (0.6% and 0.2%, respectively; P = 0.0223). Psychoactive medication was dispensed to more patients diagnosed by their primary care physicians than by their psychiatrists (85.4% and 77.3%, respectively; P = 0.0001). Fewer psychiatrists’ patients, compared with primary care physicians’ patients, received antidepressants (37.7% and 58.9%, respectively; P = 0.0001). However, more psychiatrists’ patients received stimulants (0.8% and 0.1%, respectively; P = 0.0003) and major tranquilizers (20.1% and 4.0%, respectively; P = 0.0001). In univariate analyses, female sex, depression NOS, comorbidity, and diagnosing primary care physicians were associated with increased odds of receiving psychoactive medication (Table 3, left side). In multivariate analysis, female sex, depression NOS, increasing comorbidity, and living in rural areas were independent predictors of receiving psychoactive medication.
Recommended First-Line Pharmacotherapy At the first prescribing visit, 2812 (63.6%) treated patients received an antidepressant; of these, 1413 (53.9%) were new users. Further analyses performed on the 2623 patients who received psychoactive medication and were not previous users of antidepressants showed that 1310 (49.9%) received first-line pharmacotherapy. Results of the univariate models for receiving recommended first-line pharmacotherapy (Table 3, right side) indicated that male sex , depression NOS, living in rural areas, having a diagnosing primary care physician, having the same diagnosing and prescribing physician, and receiving the medication within 31 days of diagnosis were associated with increased odds of receiving first-line pharmacotherapy. In multivariate analysis, male sex, depression NOS, having the same diagnosing and prescribing physician, and receiving the medication within 31 days of diagnosis were independently associated with increased odds of receiving recommended first-line pharmacotherapy. DiscussionWe examined the psychoactive medication claims data in Quebec for seniors with newly diagnosed depression during 2000–2002. Eighty-four percent of patients were dispensed psychoactive medication following diagnosis; of these patients, 87% received this medication shortly after diagnosis. Women were more likely than men to receive psychoactive medication, a finding which corroborates that of others (30), and may be due to either increased disease severity or greater health care–seeking behaviour by women with depression (30,31). Depression NOS increased the likelihood of receiving psychoactive medication, possibly because physicians used this diagnostic code to indicate major depression in the absence of such a code in the ICD-9 (21). By contrast, ICD-9 codes exist for both dysthymia and bipolar disorder (21). Possibly, patients with medical comorbidity were more likely to receive psychoactive medication because they had poorer quality of life and functioning, compared with patients without comorbid conditions (32). Patients residing in urban areas were less likely to receive psychoactive medication, possibly because they had easier access to psychotherapy. A sensitivity analysis performed on a subsample that excluded previous users of antidepressants supported these findings. Most late-life depression was diagnosed and treated in general practice. Greater proportions of women than men were older, diagnosed with depression NOS, diagnosed by primary care physicians, received more medication after diagnosis and within 31 days, had the same diagnosing and treating physician, and received antidepressants prior to diagnosis. Use of most psychoactive medications increased following diagnosis—especially use of antidepressants, which rose from 36% in the year prior to diagnosis to 64% in the subsequent year. By comparison, benzodiazepine use was high prior to and subsequent to diagnosis, which mirrors others’ findings (33,34) and suggests that use is long-term. Although one Canadian study found that elderly subjects with depression were more likely to be prescribed benzodiazepines than antidepressants (35), only 18% of patients in our study received benzodiazepines without antidepressants. The finding that benzodiazepine prescribing persists, despite evidence of increased risks for recurrence and relapse of depression (36), as well as for cognitive problems (35,37) and falls in elderly patients (38), should raise concerns about appropriate prescribing for the elderly. Psychiatrists possibly prescribed psychoactive medication less often than primary care physicians because they provided psychotherapy. Having the same physician diagnosing and treating the depression was the strongest predictor of receiving recommended first-line pharmacotherapy. Relational continuity (39) was more common among primary care physicians than psychiatrists (81.9% and 25.3%, respectively; P = 0.0001) and among patients dispensed psychoactive medication soon after diagnosis (78.1% and 42.0%, respectively; P = 0.0001), suggesting that long-standing relationships with primary care physicians are important to receiving appropriate pharmacotherapy. Because the Guidelines (18) were developed for the treatment of depression, we regenerated the analyses for first-line pharmacotherapy and excluded patients with dysthymia. The results corroborated our initial findings, although lower comorbidity became a significant predictor (OR 0.97; 95%CI, 0.60 to 0.92; P = 0.0275). Collectively, these findings support the notion that older patients are more likely to follow suggestions from physicians they know well. Despite having documented depressive disorders, women were less likely than men to receive first-line pharmacotherapy, even though they were more likely to receive psychoactive medication. Sex differences in care received are reported in other health areas as well (40–42), pointing to the need for further research to elucidate why guidelines and accepted standards of practice are followed less often in women’s, compared with men’s, health care. Population administrative databases provide many advantages but are not without their shortcomings. Validity of diagnoses cannot be established, although physicians tend to underdiagnose depression (10,43–45). Pharmacy claims were proxies for patient consumption and (or) physician prescribing. However, comparison of claims records to patient self-report of antidepressant use is satisfactory (kappas ~ 0.70) (46,47). Information on case severity, medication samples provided (46), and use of psychotherapy by psychologists and other allied health professionals was missing from the RAMQ databases. The generalizability of our findings may be limited to elderly patients with uncomplicated depression because we excluded patients diagnosed with other psychiatric conditions in the previous year. Notwithstanding these limitations, this study is population-based and includes seniors from all socioeconomic strata. We studied only patients with diagnosed depression because antidepressants may be prescribed for medical conditions not addressed by depression practice guidelines. Finally, by focusing on new episodes of depression and the initial medication dispensed, we evaluated receipt of first-line pharmacotherapy recommended at the time of the study. In conclusion, most seniors with newly diagnosed depression received psychoactive medication, although many did not receive recommended first-line pharmacotherapy. The finding that male sex was associated with receiving recommended first-line pharmacotherapy may increase physicians’ awareness that women are at risk for not receiving appropriate medication. The finding that benzodiazepines were dispensed to nearly one-half of treated patients may raise concerns about prescribing patterns because these medications may worsen depression or lead to adverse events. Of pivotal importance, relational continuity was the strongest predictor of receiving recommended first-line pharmacotherapy, suggesting that the patient–provider relationship is key to receiving appropriate medication. Funding and SupportFinancial support for this work was provided by the Canadian Institutes of Health Research. Dr Sewitch is supported as a research scientist of the Canadian Cancer Society through an award from the National Cancer Institute of Canada. During this study, she was supported by a postdoctoral fellowship from the Canadian Institutes of Health Research. Dr Rahme is supported by the Canadian Arthritis Society. AcknowledgmentsWe gratefully acknowledge MichPèe Paré’s contribution to the data analysis. References1. Gallo JJ, Lebowitz BD. The epidemiology of common late-life mental disorders in the community; themes for the new century. Psychiatr Serv 1999;50:1158–66. 2. Johnson J, Weissman MM, Klerman GL. Service utilization and social morbidity associated with depressive symptoms in the community. JAMA 1992;267:1478–83. 3. Crystal S, Sambamoorthi U, Walkup JT, Akincigil A. Diagnosis and treatment of depression in the elderly medicare population: predictors, disparities, and trends. JAGS 2003;51:1718–28. 4. Patten SB. Incidence of major depression in Canada. CMAJ 2000;163:714–5. 5. Cole M, Bellavance F, Mansour A. Prognosis of depression in elderly community and primary care populations: a systematic review and meta-analysis. Am J Psychiatry 1999;156:1182–9. 6. Lin EHB, Katon W, Von Korff M, Tang L, Williams JWJr, Kroenke K, and others. Effect of improving depression care on pain and functional outcomes among older adults with arthritis. A randomized controlled trial. JAMA 2003;290:2428–34. 7. Frasure-Smith N, Lespérance F. Depression and other psychological risks following myocardial infarction. Arch Gen Psychiatry 2003;60:627–36. 8. Howland RH. General health, health care utilization, and medical comorbidity in dysthymia. Int J Psychiatry Med 1993;23:211–38. 9. Demers M. Frequent users of ambulatory health care in Quebec: the case of doctor-shoppers. CMAJ 1995;153:37–42. 10. Katon W, Von Korff M, Lin E, Lipscomb P, Russo J, Wagner E, and others. Distressed high utilizers of medical care. DSM-III-R diagnoses and treatment needs. Gen Hosp Psychiatry 1990;12:355–62. 11. Callahan CM. Quality improvement research on late life depression in primary care. Med Care 2001;39:772–84. 12. Sirey J, Meyers BS, Bruce ML, Alexopoulos GS, Perlick DARP. Predictors of antidepressant prescription and early use among depressed outpatients. Am J Psychiatry 1999;156:690–6. 13. Moride Y, Galbaud Du Fort G, Monette J, Ducruet T, Boivin J-F, Champoux N, and others. Sub-optimal duration of antidepressant treatments in the older ambulatory population of Quebec: association with selected physician characteristics. JAGS 2002;50:1371. 14. Fairman KA, Drevets WC, Kreisman JJ, Teitelbaum F. Course of antidepressant treatment, drug type, and prescriber’s specialty. Psychiatr Serv 1998;49:1180–6. 15. Rojas-Fernandez C, Thomas VS, Carver D, Tonks R. Suboptimal use of antidepressants in the elderly: a population-based study in Nova Scotia. Clin Therapeut 1999;21:1937–50. 16. Unutzer J. Diagnosis and treatment of older adults with depression in primary care. Biol Psychiatry 2002;52:285–92. 17. Percudani M, Barbui C, Fortino I, Petrovich L. Antidepressant drug prescribing among elderly subjects: a population-based study. Int J Geriatr Psychiatry 2005;20:113–8. 18. Thorpe L, Whitney DK, Kutcher SP, Kennedy SH, and the CANMAT Depression Work Group. Clinical guidelines for the treatment of depressive disorders. Can J Psychiatry 2001;46(Suppl 1). 19. Institut de la statistique du Québec. Nombre de personnes inscrites et admissibles au régime d’assurance maladie du Québec, selon le sexe, le groupe d’âge et la région sociosanitaire, 2002. Québec; Gouvernement du Québec; 2002. 20. Tamblyn R, Lavoie G, Petrella L, Monette J. The use of prescription claims databases in pharmacoepidemiological research: the accuracy and comprehensiveness of the prescription claims database in Quebec. J Clin Epidemiol 1995;48:999–1009. 21. World Health Organization. International classification of diseases. 9th revision. Geneva (CH): WHO; 1975. 22. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. Washington (DC): APA; 1994. 23. Von Korff M, Wagner EH, Saunders K. A chronic disease score from automated pharmacy data. J Clin Epidemiol 1992;45:197–203. 24. Baldwin RC, Anderson D, Black S, Evans S, Jones R, Wilson K, and others. Guideline for the management of late-life depression in primary care. Int J Geriatr Psychiatry 2003;18:829–38. 25. Solai LK, Mulsant BH, Pollock BG. Selective serotonin reuptake inhibitors for late-life depression: a comparative review. Drugs Aging 2001;18:355–68. 26. Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13–22. 27. SAS Institute Inc. SAS Version 8.02. Cary (NC): SAS Inc; 2001. 28. SAS Institute Inc. The GENMOD procedure. SAS/STAT software: changes and enhancements for release 6.12. Cary (NC): SAS Inc; 1996. p 23–42. 29. Akaike H. A new look at statistical model identification. IEEE Trans Auto Control 1974;19:716–23. 30. Angst J, Gamma A, Lepine J-P, Mendlewicz J, Tylee A. Gender differences in depression. Epidemiological findings from the European DEPRES I and II studies. Eur Arch Psychiatry Clin Neurosci 2002;252:201–9. 31. Roness A, Mykletun A, Dahl AA. Help-seeking behaviour in patients with anxiety disorder and depression. Acta Psychiatr Scand 2005;111:51–8. 32. Sheikh JI, Cassidy EL, Doraiswamy PM, Salomon RM, Holland PJ, Mandel JS, and others. Efficacy, safety, and tolerability of sertraline in patients with late-life depression and comorbid medical illness. JAGS 2004;52:86–92. 33. Valenstein MA, Taylor KK, Austin K, Kales HC, McCarthy JF, Blow FC. Benzodiazepine use among depressed patients treated in mental health settings. Am J Psychiatry 2004;161:654–61. 34. Balestrieri M, Maron G, Marini M, Sessa E, Gelatti U, Donato F. Mental disorders associated with benzodiazepine use among older primary care attenders. A regional survey. Soc Psychiatry Psychiat Epidemiol 2005;40:315. 35. Hogan DB, Maxwell CJ, Fung TS, Ebly EM. Canadian Study of Health and Aging. Prevalence and potential consequences of benzodiazepine use in senior citizens: results from the Canadian Study of Health and Aging. Can J Clin Pharmacol 2005;10:72–7. 36. Melfi CA, Chawla AJ, Croghan TW, Hanna MP, Kennedy S, Sredl K. The effects of adherence to antidepressant treatment guidelines on relapse and recurrence of depression. Arch Gen Psychiatry 1998;55:1128–32. 37. Paterniti S, Dufouil C, Alperovitch A. Long-term benzodizepine use and cognitive decline in the elderly: the epidemiology of vascular aging study. J Clin Psychopharmacol 2002;22:285–93. 38. Tamblyn R, Abrahamowicz M, du Berger R, McLeod P, Bartlett G. A 5-year prospective assessment of the risk associated with individual benzodiazepines and doses in new elderly users. JAGS 2005;53:233–41. 39. Haggerty JL, Reid RJ, Freeman GK, Starfield BH, Adair CE, McKendry R. Continuity of care: a multidisciplinary review. BMJ 2003;327:1219–21. 40. Turner B, Myers RE, Hyslop T, Hauck WW, Weinberg D, Brigham T, and others. Physician and patient factors associated with ordering a colon evaluation after a positive fecal occult blood test. J Gen Int Med 2003;18:357–63. 41. Daly SC, Roger VL, Leibson C. Cardiology services after stress testing: are there sex differences? A population-based study. J Clin Epidemiol 2000;53:661–8. 42. Mocroft A, Gill MJ, Davidson W, Phillips AN. Are there gender differences in starting protease inhibitors. HAART and disease progression despite equal access to care? JAIDS 2000;24:475–82. 43. Montano CB. Recognition and treatment of depression in a primary care setting. J Clin Psychiatry 1994;55:18–34. 44. Rost K, Smith GR, Matthews DB, Guise B. The deliberate misdiagnosis of major depression in primary care. Arch Fam Med 1994;3:337. 45. Broadhead WE. Misdiagnosis of depression. Physicians contribute to the stigmatization of mental illness. Arch Fam Med 1994;3:319–20. 46. Kwon A, Bungay KM, Pei Y, Rogers WH, Wilson IB, Zhou Q, and others. Antidepressant use. Concordance between self-report and claims records. Med Care 2003;41:368–74. 47. Saunders K, Simon G, Bush T, Grothaus L. Assessing the feasibility of using computerized pharmacy claims refill data to monitor antidepressant treatment on a population basis: a comparison of automated and self-report data. J Clin Epidemiol 1998;51:883–90. Author(s)Manuscript received July 2005, revised, and accepted January 2006. 1. Assistant Professor, McGill University, Montreal, Quebec. 2. Professor, Université de Montréal, Montreal, Quebec. 3. Assistant Professor, Université de Montréal, Montreal, Quebec. 4. Faculty Lecturer, Université de Montréal, Montreal, Quebec; President, Association des médecins psychiatres du Québec, Montreal, Quebec. Address for correspondence: Dr MJ Sewitch, 1650 Cedar Avenue, Room L10-409, Montreal General Hospital, Montreal, QC H3G 1A4 e-mail: Maida.Sewitch@mcgill.ca
1 | 2
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||