stepped care (chile)

Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/10835736

Treatingdepressioninprimarycareinlow-incomewomeninSantiago,Chile:arandomisedcontrolledtrialARTICLEinTHELANCETMARCH2003ImpactFactor:45.22DOI:10.1016/S0140-6736(03)12825-5Source:PubMed

CITATIONS133

DOWNLOADS96

VIEWS108

7AUTHORS,INCLUDING:

GracielaRojasUniversityofChile83PUBLICATIONS632CITATIONS

SEEPROFILE

RosemarieFritschUniversityoftheAndes(Chile)110PUBLICATIONS956CITATIONS

SEEPROFILE

JorgeGaeteLondonSchoolofHygieneandTropicalMe48PUBLICATIONS336CITATIONS

SEEPROFILE

Availablefrom:RosemarieFritschRetrievedon:11July2015

For personal use. Only reproduce with permission from The Lancet Publishing Group.

ARTICLES

THE LANCET Vol 361 March 22, 2003 www.thelancet.com 995

Summary

Background Depression in women is one of the commonestproblems encountered in primary care. We aimed tocompare the effectiveness of a stepped-care programmewith usual care in primary-care management of depressionin low-income women in Santiago, Chile.

Methods In a randomised controlled trial, in three primary-care clinics in Chile, 240 adult female primary-care patientswith major depression were allocated stepped care or usualcare. Stepped care was a 3-month, multicomponentintervention led by a non-medical health worker, whichincluded a psychoeducational group intervention, structuredand systematic follow-up, and drug treatment for patientswith severe depression. Data were analysed on an intention-to-treat basis. The primary outcome measure was theHamilton depression rating scale (HDRS) administered atbaseline and at 3 and 6 months after randomisation.

Findings About 90% of randomised patients completedoutcome assessments. There was a substantial between-group difference in all outcome measures in favour of thestepped-care programme. The adjusted difference in meanHDRS score between the groups was 889 (95% CI 1115to 676; p


covered included information on symptoms and causesof depression, available treatment options, schedulingpositive activities, problem-solving techniques, and basiccognitive and relapse-prevention techniques. Patientswere given a manual with detailed information of thecontents of each session accompanied by examples andexercises. Group leaders were social workers and nurseswho received 12 h of training and 8 h of supervisionfrom the principal investigators. Most group leaderswere employed in local primary-care clinics. Patientswere discouraged from contacting the group leaderoutside the sessions unless essential.

Patients with severe depression (HDRS score >19) atbaseline or persistent depression (HDRS score >12)after 6 weeks of group treatment were referred back totheir primary-care physician for a structuredpharmacotherapy programme.24 Primary-care physiciansreceived 4 h of training to enable them to deliver a briefpharmacotherapy protocol that included structuredassessment at initial and follow-up visits and the use of astandard medication algorithm to ensure adequate doseand duration of treatment (fluoxetine, amitriptyline, orimipramine). Group leaders monitored medicationadherence and attendance at follow-up visits for patientsreceiving pharmacotherapy (figure 1).

Patients assigned usual care received all servicesnormally available in the primary-care clinic, includingantidepressant medication or referral for specialtytreatment. Before the initiation of the study, primary-care physicians in the control group received guidelineson how to treat depression in primary care. No services normally available were restricted or withheld,and primary-care physicians received no informationfrom study workers about patients in the usual-caregroup.

All participants were invited to attend the primary-care clinics for outcome assessments at 3 and 6 monthsafter randomisation. Follow-up interviews were done byan independent clinician blinded to treatmentassignment. Patients attending outcome assessmentsreceived a small payment to cover travel expenses.

ARTICLES

996 THE LANCET Vol 361 March 22, 2003 www.thelancet.com

2000, and November, 2001. The clinics were represent-ative of Santiago primary-care clinics in terms ofresources and of clinical and sociodemographiccharacteristics of patients.17 We used a two-stagescreening process to identify female primary-carepatients aged 1870 years with current major depressiveillness. Consecutive female patients were approachedwhile they were waiting for consultation. All eligible andconsenting patients were asked to complete the generalhealth questionnaire (GHQ-12). All women with a scoreof 5 or more were asked to return for another GHQ-12assessment 2 weeks later. Those scoring 5 or more at thesecond screening were invited for a baseline assessmentwithin 1 week. The double screening was part of thestepped-care strategy to direct resources toward patientswith persistent depression.

Ethics approval was granted by Comite Etica,Hospital Clinico, University of Chile. Patients gavewritten informed consent before being invited toparticipate.

ProceduresAt baseline, a clinician administered three assessments:the mini international neuropsychiatric interview(MINI)18 to ascertain a Diagnostic and Statistical Manualof Mental DisordersFourth Edition (DSM-IV) diagnosisof major depression, manic or psychotic episode, oralcohol abuse; the Hamilton depression rating scale(HDRS);19 and the short form 36 questionnaire (SF-36).HDRS, the primary outcome measure, has been shownto give reliable and valid results in primary-carepatients,20 as has the Spanish-language version.21 SF-36is used to assess functional impairment across a range ofmental and physical domains.22 Four aspects of SF-36were selected as secondary outcomes: mental health,emotional role, social functioning, and vitality (selectedbecause of responsiveness to change in depression). Inprimary-care populations similar to ours it has showngood reliability and sensitivity to change with severity ofdepression. A Spanish-language version has been usedsuccessfully in studies of depression in Spanish-speakingprimary-care patients.23

All patients with current DSM-IV major depressionwere eligible. Patients with current psychotic symptoms,serious suicidal risk, history of mania, or current alcoholabuse were excluded and referred back to their primary-care physician. Patients who had had a psychiatricconsultation or admission to hospital in the 3 monthsbefore the interview were also excluded.

Patients were randomly assigned the stepped-careimprovement programme or usual care. Randomisationwas stratified by clinic and done in blocks of 20 by use ofcomputer-generated random numbers; individuals whorecruited patients were neither involved in nor aware ofthe procedure used to generate allocations. Allocationswere kept in numbered sealed envelopes in each clinic,and were opened by an individual who had not recruitedpatients. The method of randomisation was chosen toobtain groups of equal size.

The stepped-care improvement programme was amulticomponent programme consisting of a structuredpsychoeducational group, systematic monitoring ofclinical progress, and a structured pharmacotherapyprogramme for patients with severe or persistentdepression. Figure 1 shows an overview of theprogramme. The psychoeducational intervention groupconsisted of seven weekly sessions and two boostersessions at weeks 9 and 12; each session lasted 75 min.Each group included about 20 participants. Topics

Baseline assessment

HDRS score 19 HDRS score >19

Booster sessionsweeks 9 and 12

HDRS score 12 HDRS score >12

Psychoeducational group7 weekly sessions

Psychoeducational groupand assessment

for pharmacotherapy

Refer for primary-carephysician reassessment:initiate pharmacotherapy

oradjust pharmacotherapy

6-week reassessment

Figure 1: The stepped-care intervention programme


ARTICLES


Statistical analysisOn the basis of data from other primary-care studies, weanticipated that follow-up HDRS scores would have anSD of about 7 points. Thus, a sample size of 120 patientsper group would have 80% power for detection of adifference of 25 points in mean HDRS score betweengroups with two-sided significance of 5%.

Data were analysed in accordance with CONSORTguidelines; between-group comparisons were byintention-to-treat. We obtained descriptive statistics forthe primary outcome measure (HDRS score) as acontinuous and categorical (proportion of patientsimproved [defined as a 50% score reduction] andrecovered [score


analysis after adjusting for the large differences in theproportions prescribed antidepressants in the two groups.The estimate, 95% CIs, and p value for HDRS score wasnot much changed by either this adjustment or bycontrolling for the small baseline imbalance in chronicmedical conditions. Likewise, a sensitivity analysis hadalmost no effect on the results, which was as expected onthe basis of the data presented in figure 3. To aidinterpretation of the primary analyses in table 2,descriptive statistics for the binary versions of HDRSscores are presented in table 3. For example, theproportion of women who recovered (HDRS


ARTICLES


the clinicians who did outcome assessments learned oftreatment allocation. However, we obtained similarresults using a self-reported measure, the SF-36 mentalhealth dimension. It is therefore unlikely that anassessment bias had much effect on our results.

Patients in the stepped-care programme receivedseveral treatment components (patients education,behavioural activation, problem-solving techniques,systematic and structured follow-up, and/orpharmacotherapy). We did not aim to establish theeffectiveness of isolated components of treatment, butrather to assess a programme that included componentsshown to be effective, available locally, and that could beused in an efficient stepped-care programme. In otherwords, our programme was intended to be as close aspossible to what should be an adequate, efficient, andmore importantly, feasible treatment programme in ourparticular setting. Even though there were largedifferences in the proportion of people in each group whowere prescribed antidepressants, we think it is unlikelythis could account for the significant differences inoutcomes between the groups since adjusting our resultsfor medication did not alter our main findings.

We did not design the programme to introduce radicalchanges to established practices. We aimed to improveexisting care with structured protocols and rationalise theuse of available resources. Often, innovative programmesare introduced that involve changes impossible toassimilate in the day-to-day delivery of care. The maininnovative element was role enhancement for the non-medical group leaders, most of whom were available inthe clinics and were often closely connected to localneighbourhoods. These workers were given the time andtraining to deliver the group psychological interventioneffectively, monitor treatment progress, and act asadvocates or care managers. They also acted as brokersbetween patients and doctors without interfering withdoctors work. On the contrary, doctors appreciated thehelp and information. We decided to concentrate ourintervention in a high-risk group, poor women, so thatthe psychoeducational group intervention could focus ontopics relevant for that group. This strategy seems to haveworked because attendance rates and overall satisfactionwith the programme were reasonably good.

Our results have some limitations. First, research hasshown that benefits gained early during treatment tend todecline after 1 year or longer.11,26,27 Unfortunately, wewere unable to do a 1-year assessment, but our 6-monthresults compare favourably with those of otherprogrammes.1113,26 Second, the usual-care group didworse than in other studies from developed countries, anunsurprising finding in view of the deficiencies in primaryhealth care and the characteristics of patients in theChilean sample. Third, our intervention was designed forease of implementation in other primary-care settings,but we cannot be certain that our results are generallyapplicable. Finally, our psychological interventionincluded nine group sessions, which could be thoughtunrealistically intense in some settings. However,psychological interventions of similar intensity hadalready been tried with success in primary-care settings inChile and the costs were acceptable to services andpatients. Our main aim was to show that organised,improved treatment, as a first step, could effectively makea difference in a setting of socioeconomic disadvantage.Future interventions should test the effectiveness of less-intensive and costly psychological interventions.

Clinics and patients were representative of primarycare in Santiago, Chile,5,17 with high rates of morbidity,

severe resource limitations, and socioeconomicallydeprived populations. These factors are common barriersto effective care in most public-sector clinics indeveloping countries. Thus, we think our results could beapplicable in similar settings to ours, but also that similarstrategies can be equally cost effective to managedepression or other chronic illnesses in more advantagedor organised practice settings. Contrary to someexpectations, developed countries can learn much fromdeveloping countries in how best to rationalise resourceswhen working with deprived populations under tightbudgetary conditions.

The combination of high rates of depressive illness,poverty, and scarce resources can easily induce nihilismin physicians and policy-makers. Our findings shouldoffer hope that modest interventions can have asubstantial effect on depressive symptoms and functionalimpairment. The benefits of our programme exceed thoseseen with similar interventions in settings with less-deprived patients and more treatment resources. Unlikeresults from many other successful treatment studies, ourresults were well received by local policy-makers whohave now launched a national pilot programme for thetreatment of depression in primary care based on thisintervention. Socially disadvantaged patients, especiallyin the developing world, might have the most to gainfrom modest investments in organised treatment ofdepression.

ContributorsR Araya and G Simon wrote the grant proposal. R Araya and T J Petersanalysed data and wrote the final manuscript. G Rojas was the studycoordinator and obtained data, assessed patients, and editedmanuscripts. R Fritsch obtained data, assessed patients, entered andanalysed data, and edited early versions of the manuscript. J Gaete wasan independent outcome assessor and revised manuscripts. M Rojas wasthe site coordinator and collected data and edited manuscripts. G Simonprovided advice on fieldwork and edited all versions of the manuscript.

Conflict of interest statementR Araya received payment from Wyeth for a workshop. G Rojas receivedpayment from Wyeth and Servier and R Fritsch received payment fromWyeth for participation in clinical trials. G Simon has received researchgrants from Eli Lilly and Solvay Pharmaceuticals. None of the otherauthors has declared any conflict of interest.

AcknowledgmentsWe thank the staff from the participating primary-care clinics. We alsothank all the members of our team, in particular, C Valera, N Parra, M E Romero, E Lebrini, P Gonzalez, and J Espinola. Our thanks to Prof Glyn Lewis who read the manuscript and made useful comments.We thank the US National Institute of Mental Health for funding thisproject.

References1 Murray C, Lopez A. Alternative projections of mortality and

disability by cause 19002020: Global Burden of Disease Study.Lancet 1997; 349: 1498504.

2 Patel V, Araya R, Ludemir A, Todd C, Lima M. Women, povertyand common mental disorders in four restructuring societies. Soc Sci Med 1999; 49: 146171.

3 Jenkins R, Lewis G, Bebbington P, et al. The national psychiatricmorbidity surveys of Great Britain: initial findings from the householdsurvey. Psychol Med 1997; 27: 77590.

4 Araya R, Rojas G, Fritsch R, Acuna J, Lewis G. Common mentaldisorders in Santiago, Chile: prevalence and socio-demographiccorrelates. Br J Psychiatry 2001; 178: 22833.

5 Ustun B, Sartorius N. Mental illness in general health care: aninternational study. London: John Wiley and Sons, 1995.

6 Araya R, Lewis G, Rojas G, Mann A. Patient knows best: detectionof common mental disorders in Santiago, Chile: cross-sectional study.BMJ 2001; 322: 7981.

7 Simon G, von Korff M, Barlow W. Health care costs of primary carepatients with recognized depression. Arch Gen Psychiatry 1995; 52:85056.


ARTICLES

1000 THE LANCET Vol 361 March 22, 2003 www.thelancet.com

8 Schulberg H, Block M, Madonia M, et al. The usual care of majordepression in primary care practice. J Arch Fam Med 1997; 6:33439.

9 Tylee A, Freeling P. The recognition, diagnosis andacknowledgement of depressive disorders by general practitioners. In:Paykel E, Herbst K, eds. Depression: an integrative approach.London: Heinemann, 1989: 21631.

10 Mynors-Wallis LM, Gath DH, Day A, Baker F. Randomisedcontrolled trial of problem solving treatment, antidepressantmedication, and combined treatment for major depression in primarycare. BMJ 2000; 320: 2630.

11 Ward E, King M, Lloyd M, et al. Randomised controlled trial onnon-directive counselling, cognitive-behaviour therapy, and usualgeneral practitioner care for patients with depression. I: clinicaleffectiveness. BMJ 2000; 321: 138388.

12 Chilvers C, Dewey M, Fielding K, et al. Antidepressant drugs andgeneric counselling for treatment of major depression in primary care:randomised trial with patient preference arms. BMJ 2001; 322:77275.

13 Malt UF, Robak H, Madsbu HP, Bakke O, Loeb M. The Norwegiannaturalistic treatment study of depression in general practice(NORDEP) I: randomised double blind study. BMJ 1999; 318:118084.

14 Katon W, Robinson P, von Korff M, et al. A multifacetedintervention to improve treatment of depression in primary care. ArchGen Psychiatry 1996; 53: 92432.

15 Mynors-Wallis LM, Davies I, Gray A, Barbour F, Gath D. Arandomised controlled trial and cost analysis of problem-solvingtreatment for emotional disorders given by community nurses inprimary care. Br J Psychiatry 1997; 170: 11319.

16 Simon G, von Korff M, Rutter C, Wagner EH. Randomised trial ofmonitoring, feedback, and management of care by telephone to

improve treatment of depression in primary care. BMJ 2000; 320:55054.

17 Araya R. Common mental disorders and detection by primary carephysicians. PhD thesis, University of London, 2000.

18 Sheehan DV, LeCrubier Y, Janrs J, et al. Mini internationalneuropsychiatric interview (MINI). Florida: University of SouthFlorida Institute for Research on Psychiatry and INSERM-HpitalPiti Salptrire, 1994.

19 Hamilton M. A rating scale for depression. J Neurol NeurosurgPsychiatry 1960; 23: 5662.

20 Simon G, VonKorff M, Revicki D. Telephone assessment ofdepression severity. J Psychiatr Res 1993; 27: 24752.

21 Ramos-Brieva J, Cordero-Villafalia A. A new validation of theHamilton rating scale for depression. J Psychiatr Res 1988; 22: 2128.

22 Ware J, Kosinski M, Keller SD. SF-36 physical and mentalcomponent summary scales: a users manual. Boston: The HealthInstitute, New England Medical Center, 1993.

23 Williams J, Kerber C, Mulrow C, Medina-Aguilar ME. Depressivedisorders in primary care: Prevalence, functional disability, andidentification. J Gen Int Med 1995; 10: 712.

24 Katon W, von Korff M, Lin E, et al. Collaborative management toachieve treatment guidelines: the impact on depression in primarycare. JAMA 1995; 273: 102631.

25 Armitage P, Berry G, Matthews JNS. Statistical methods in medicalresearch, 4th edn. Oxford: Blackwell Science, 2002.

26 Dowrick C, Dunn G, Ayuso-Mateos JL, et al, and the Outcomes of Depression International Network (ODIN) Group. Problemsolving treatment and group psychoeducation for depression:multicentre randomised controlled trial. BMJ 2000; 321: 145054.

27 Sherbourne C, Wells K, Duan N, et al. Long-term effectiveness ofdisseminating quality improvement for depression in primary care.Arch Gen Psychiatry 2001; 58: 696703.

Wernickes encephalopathy

Markus Ploner, Alfons Schnitzler

Clinical picture

Department of Neurology, Heinrich-Heine-University, 40225 Dsseldorf, Germany (M Ploner MD, A Schnitzler MD)

A 36-year-old man was admitted to the emergency wardin a global confusional state. Apart from the patientsname and age no further information on the history was obtainable. Neurological examination revealedintermittent upbeat nystagmus, convergent strabismus,horizontal gaze palsy, and a severe ataxia of stance andgait. Laboratory findings showed an elevation ofgamma-glutamyl transpeptidase, serum amylase andserum lipase, and a macrocytic anaemia. Drug andalcohol testing was negative. Based on the classical triad of clinical featuresophthalmoplegia, ataxia, andconfusionWernickes encephalopathy was diagnosedand thiamine was administered immediately. A fluid-attenuated inversion recovery (FLAIR) magneticresonance-scan showed hyperintense signals around the third ventricle, in the fornix, and around the fourthventricle, the hypothalamus, and the mamillary bodies(figure): findings characteristic of Wernickes encephalopa-thy. Cerebrospinal fluid showed a small elevation ofprotein content. Within the first days ocular palsies nearlycompletely recovered while ataxia and confusionremained unchanged. A history of chronic alcohol abusewas confirmed.

stepped care (chile)

Documents