Journal of Psychosomatic Research 72 (2012) 419–421

Contents lists available at SciVerse ScienceDirect

Journal of Psychosomatic Research

Is co-morbid depression adequately treated in patients repeatedly referred tospecialist medical services with symptoms of a medical condition?

Michael Sharpe a,⁎, Christopher Burton b, Aarti Sawhney c, Kelly McGorm b, David Weller b

a Psychological Medicine Research, Department of Psychiatry, University of Oxford, UKb Centre for Population Health Sciences, University of Edinburgh, UKc Psychological Medicine Research, School of Molecular and Clinical Medicine, University of Edinburgh, UK

⁎ Corresponding author at: Psychological Medicine Reatry, University of Oxford, UK. Tel.: +44 1865 226397;

E-mail address: michael.sharpe@psych.ox.ac.uk (M.

0022-3999/$ – see front matter © 2012 Elsevier Inc. Alldoi:10.1016/j.jpsychores.2012.03.005

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 4 November 2011Received in revised form 6 March 2012Accepted 6 March 2012

Keywords:DepressionReferralTreatment

Objective: Patients with a medical condition and co-morbid depression have more symptoms and use moremedical services. We aimed to determine the prevalence of depression and the adequacy of its treatmentin patients who had been repeatedly referred from primary to specialist medical care for the assessment ofa medical condition.Methods: All patientswhohad at least three referrals tomedical and surgical specialists for an assessment of symp-toms attributed to a medical condition, over a five year period from five primary care practices in Edinburgh, UKwere identified using a referral database and review of records. Participants were sent a questionnaire whichincluded the PHQ-9 depression scale and additional questions about depression during the preceding 5 years.

Details of treatment for depression were obtained from primary care records.Results: Questionnaires were sent to 230 patients and returned by 162 (70.4%). Forty-one (25.3%) had a PHQ-9score of 10 or more and hence probable current depressive disorder. An additional 36 (22.2%) reported depres-sion in the previous 5 years. Only eight (19.5%) of those reporting current depression and 20 (26%) of the 77patients reporting previous depression had received minimally adequate treatment for it.Conclusion: Whilst we know that patients with medical conditions are often depressed and that such co-morbiddepression is often undertreated, we have found that it is undertreated even in patients repeatedly referred tomedical specialists. Better assessment and management of depression in such patients could both improvepatients' quality of life and reduce the cost of care.

© 2012 Elsevier Inc. All rights reserved.

Introduction

Patients who have a medical condition complicated by depression,so called co-morbid depression, are known to have more physicalsymptoms and to be higher users of medical services than patientswith the medical condition but without depression [1,2]. An impor-tant and potentially expensive aspect of high service use is the re-peated referral of patients to hospital medical specialists by theirgeneral practitioners.

The effective treatment of co-morbid depression in such patientsis important as it may both improve the patient's symptoms and re-duce the need for the repeated referral for the assessment of these.However, we do not know how effectively co-morbid depression iscurrently treated in these repeatedly referred patients.

We have previously reported on a representative primary caresample of repeatedly referred patients and specifically on depressionin the subsample of patients with medically unexplained symptoms

search, Department of Psychi-fax: +44 1865 793101.Sharpe).

rights reserved.

(MUS) [3,4]. Here we studied the same repeatedly referred sampleto address the question of depression and the adequacy of its treat-ment in the subsample of patients with a medical condition (medical-ly explained symptoms).

The aim of this analysis was to determine: (a) the prevalence ofcurrent and previous co-morbid depression in this sample and(b) the adequacy of treatment received for that depression.

Methods

Identification of patients

As part of the larger study we identified all patients who had re-ceived at least three new referrals tomedical and surgical specialist ser-vices all of whichwere for the assessment of amedical condition, over afive year period. The methods are described in detail elsewhere [5].Briefly, the initial study sample was of all patients registered with fiveprimary care practices, representing 30 general practitioners (GPs) inEdinburgh, UK and serving practice populations totaling to 39,562 pa-tients. The practices were all multi-partner practices used for teaching.They were chosen to represent a range of socio-economic areas. The

Table 1Number and percentage patients with current or past depression who receivedminimallyadequate treatment for depression

Currentdepressiona(n=41)

Past depressionb

(n=36)Current or pastdepressionc(n=77)

Treatmentd N % (95% CI) N % (95% CI) N % (95% CI)

Currente 8 19.5 (7.4 to 31.6) 6 16.7 (4.5 to 28.8) 14 18.2 (9.6 to 26.8)Pastf 5 12.2 (2.2 to 22.2) 1 2.8 (0 to 8.1) 6 7.8 (1.8 to 13.8)None 28 68.3 (54 to 82.5) 29 80.6 (67.6 to 93.5) 57 74.0 (64.2 to 83.8)

a Current score of 10 or more on the PHQ-9.b Self-reported depression at some time over the previous 5 years but not currently.c Total number of depressed patients over 5 years (current or past depression).d Minimally effective drug or non-drug current depression treatments (see text).e Treatment recorded in the last 60 days (drugs) or last year (psychological treatment).f Treatment given at some time in past five years, but not currently (as defined above).

420 M. Sharpe et al. / Journal of Psychosomatic Research 72 (2012) 419–421

number and destination of referrals of patients from the participatingpractices were obtained using an NHS database of hospital activity. Allpatients who had been referred to one of a number of medical special-ities three ormore times in the previous 5 years andwhosemost recentreferral had been within the last year were identified. The correspon-dence from the specialist to the primary care doctor for each of thesepatients was reviewed and the patients who, in the specialists' opinion,had symptoms attributable to a medical condition on every occasionthey were referred were selected for the study.

Assessment of depression

The sample of patients selected by the above procedure was sent aconsent form and a questionnaire by post. Current depression wasassessed using the nine-item self-report Patient Health Questionnaire(PHQ-9) [6]. We used the recommended PHQ-9 total score cut-off of10 or more to indicate a probable depressive disorder. This cut-off hasbeen supported in a systematic review as valid for detecting majordepression in primary care patients and in the medically ill [7]. Fol-lowing more recent suggestions that a higher cut-off score of 11may be preferable [8], we also assessed the effect on the findings ofusing this cut-off. Past depression, defined as having experiencedlow mood or loss of interest and pleasure for more than 2 weeks atany time during the preceding 5 years, was also assessed in thequestionnaire.

Adequacy of depression treatment

We obtained details of treatment for depression (both drug andpsychological) from the primary care records. For each patient wesought to determine if minimally adequate treatment had beengiven. For drug treatment we required a record of at least one pre-scription of the recommended minimum effective dose of an antide-pressant [9]. For psychological treatment we required a record of atleast two sessions with a mental health professional of any discipline.Treatment was categorized as current if given within the previous60 days for a prescription (judged to be the reasonable maximum pe-riod for which a prescription may be given) and in the past year forpsychological treatment (judged to be the reasonable maximum in-terval between treatment sessions). Treatment before these timewindows but within the last 5 years was coded as past treatment.

Analysis

We first described the derivation of the sample and their charac-teristics. We then determined the number of patients in the samplewho had current depression and the number who had past (but notcurrent) depression to determine the total number who had been de-pressed over the previous 5 years. We then calculated how many ofthese patients were currently receiving minimally effective therapy(drug and psychological) and how many had received this duringthe previous 5 years.

Results

Of the total registered population of 39,562 patients, we identified 249 who met ourcriteria for repeated referrals, all of which were for symptoms of a medical condition(medically explained symptoms). Nineteen of these patients were no longer registered inthe practice or were regarded by their GP as unsuitable to be approached for participationin a research study. Questionnaires were sent to the remaining 230 patients and returnedby 162 (70.4%).

The mean age of the final sample on whom we had complete data was 50.8 years(s.d. 10.3). 96 (59%) patients were female and 90 (56%) were in paid employment.

Forty-one (25.3%) of the sample had a PHQ-9 score of 10 or more and hence had aprobable depressive disorder (if a cut-off score of 11 was used it was 20%). An addi-tional 36 patients reported an episode of past depression in the previous 5 years.Hence a total of 77 patients (47.5%) reported having suffered from depression atsome time in the previous 5 years.

Only eight (19.5%) of those reporting current depression were receiving minimallyadequate treatment as we defined it (if depression was defined using a PHQ-9 cut-off of11, it was 21%). Only two of these eight patients were receiving psychological treatment(one of these was also receiving antidepressant drug treatment) and the remainderwere receiving antidepressant drugs only. The proportion of patientswith past depressionwho had received such minimally adequate treatment was similar.

Overall only 20 (26%) of the 77 patients who reported depression at any time duringthe previous 5 years could be regarded as having received minimally effective treatmentfor it (see Table 1).

Discussion

We found that a quarter of the patients who had been repeatedlyreferred to hospital for the assessment of the symptoms of a medicalcondition over a five-year period had evidence of current co-morbiddepressive disorder as defined on the PHQ-9 and that almost halfreported an episode of depression at some time during the previous5 years. It was a striking finding that three quarters of these patientshad not received even minimally adequate treatment for depressiondespite their repeated referral to medical services.

We know that depression that is co-morbid with a medical con-dition often goes unrecognized and untreated [10,8]. However thefinding of such a low rate of treatment for co-morbid depression inpatients who have been repeatedly referred from primary to sec-ondary medical care for the assessment of symptoms of a medicalcondition is novel. There is little published evidence about such repeat-edly referred patients, despite the costs they incur. The prevalence ofdepression we found in this population was however similar to thatpreviously reported in a sample of patients who were frequent pre-senters to primary care [11].

Why is the rate of effective treatment for depression so low inthese patients? Whilst this particular study cannot answer this ques-tion other research suggests that there are likely to be multiple rea-sons including poor recognition in both primary and secondarymedical care, low acceptance of treatment by the patients and failureto prescribe and monitor treatment by the patients' primary and sec-ondary care doctors [12]. Previous research has reported that evenamongst patients who are enthusiastic about getting treatment fordepression, the presence of physical problems appears to reduce thelikelihood that depression will be treated because the medical condi-tion competes with depression for attention, even over multiple visitsto primary care [13]. We also know that even if recognized, depressionin the context of a medical condition may be regarded by primary caredoctors as an understandable reaction to the medical condition, ratherthan as an illness to be treated in its own right [14]. Nonetheless the ob-servation that the rate of treatment was so low in a sample of patientswho had been repeatedly assessed for referral in primary care and sim-ilarly repeatedly assessed by medical specialists in secondary care, wasespecially surprising.

There is some evidence that treating depression in such patientsmay improve their somatic symptoms and functioning [15], althoughit is not yet known if improved treatment of depression would reduce

421M. Sharpe et al. / Journal of Psychosomatic Research 72 (2012) 419–421

the number of hospital referrals, and consequently the cost of care.This is however a plausible hypothesis as more effective treatmentof depression could reduce the patient's physical symptoms and re-duce the need for repeated referral to medical specialists. A systemdesigned to achieve this must however do more than simply screenfor depression; screening alone is unlikely to improve patient outcomesin the absence of a system for treating the identified cases [16]. One ex-ample of such a treatment system is the collaborative care model inwhich a supervised case manager ensures optimal guideline-basedcare [17]. Hence we envisage a system that: (a) tells the primarycare doctor when a patient meets a threshold for frequency of refer-ral; (b) ensures that these patients are assessed for depression and(c) implements treatment guidelines such as those reported byNICE using a system such as collaborative care [18].

There were limitations to this study: first, the sample was from asmall number of primary care practices in one UK city, although thepractices did serve populations with diverse socioeconomic back-grounds. Second, not all patients returned a questionnaire, althoughthe response rate of 70% would be generally regarded as adequate.Third, depression was assessed by questionnaire rather than clinicalinterview. The measure and cut-off we used have however been val-idated in primary care populations including patients with chronicmedical illness against interview identified major depression andthe findings were similar when we used a higher cut-off of 11. Itmay be argued that simply exceeding a score on a questionnairedoes not necessarily indicate a diagnosis of depressive disorder ofmoderate severity or a need for treatment; for example an interviewmay reveal that the symptoms may be better attributed to a transientadjustment disorder; the validation studies make it unlikely that thatis the explanation in many cases however. The assessment of previ-ous depression which we based on a retrospective rating of the twomain symptoms of depression, asked after the patient had completedPHQ-9, is likely to be less accurate than that of current depression.However, the retrospectively administered PHQ-9 has been validatedas ameasure of past depression [19]. Finally, we set a very low thresholdfor minimally adequate treatment; more stringent criteria (such as tri-als of more than one antidepressant agent at full dose or delivery ofan adequate number of sessions of a competently delivered evidencebased psychological treatment [18]) would have resulted in an evenlower rate of adequate treatment.

Conclusions

The low rate of even minimal treatment for depression we foundin medically ill patients with (questionnaire identified) co-morbiddepression who had been repeatedly referred to medical specialistsreflects a focus by both doctor and patients on the management ofthe co-existing medical condition. A system of identifying repeatedlyreferred patients in primary care, assessing them for co-morbid de-pression and providing effective treatment, has the potential to im-prove patients' outcomes whilst also reducing the cost of care.

Conflict of interest

The authors have no competing interests to report.

Acknowledgments

Wewish to thank the staff of the Information and Statistics Division ofthe Scottish health Department (ISD) for their collaboration and the prac-tices andpatientswho tookpart in this study. Thisworkwas supportedbythe Chief Scientist Office of the Scottish Government Health Directorate[CZH/4/37]; the funder had no involvement in the conduct or reportingof the study.

All authors contributed to the conception and design of the studyand/or interpretation of the data and drafting of the report and allhave approved the final manuscript.

References

[1] Moussavi S, Chatterji S, Verdes E, Tandon A, Patel V, Ustun B. Depression, chronicdiseases, and decrements in health: results from theWorld Health Surveys. Lancet2007;370:851–8.

[2] Katon W. The impact of major depression on chronic medical illness. Gen HospPsychiatry 1996;18:215–9.

[3] McGorm K, Burton C, Weller D, Murray G, Sharpe M. Patients repeatedly referredto secondary care with symptoms unexplained by organic disease: prevalence,characteristics and referral pattern. Fam Pract 2010;27:479–86.

[4] Burton C, McGorm K, Weller D, Sharpe M. Depression and anxiety in patientsrepeatedly referred to secondary care with medically unexplained symptoms: acase–control study. Psychol Med 2011;41:555–63.

[5] Smith BJ, McGorm KJ, Weller D, Burton C, Sharpe M. The identification in primarycare of patients who have been repeatedly referred to hospital for medicallyunexplained symptoms: a pilot study. J Psychosom Res 2009;67:207–11.

[6] Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severitymeasure. J Gen Intern Med 2001;16:606–13.

[7] Gilbody S, Richards D, Brealey S, Hewitt C. Screening for depression in medicalsettings with the Patient Health Questionnaire (PHQ): a diagnostic meta-analysis. J Gen Int Med 2007;22:1596–602.

[8] Kendrick T, Dowrick C, McBride A, Howe A, Clarke P, Maisey S, et al. Managementof depression in UK general practice in relation to scores on depression severityquestionnaires: analysis of medical record data. BMJ 2009;338:b750.

[9] British National Formulary. London: BMJ Publishing Group; 2006.[10] Cepoiu M, McCusker J, Cole MG, Sewitch M, Belzile E, Ciampi A. Recognition of

depression by non-psychiatric physicians—a systematic literature review andmeta-analysis. J Gen Intern Med 2008;23:25–36.

[11] Katon W, Von Korff M, Lin EH, Lipscomb P, Russo JE, Wagner E, et al. Distressedhigh utilizers of medical care. DSM-III-R diagnoses and treatment needs. GenHosp Psychiatry, 12; 1990. p. 355–62.

[12] Katon W, Sullivan MD. Depression and chronic medical illness. J Clin Psychiatry1990;51:3–11.

[13] Nutting PA, Rost K, Smith J, Werner JJ, Elliot C. Competing demands from physicalproblems: effect on initiating and completing depression care over 6 months.Arch Fam Med 2000;9:1059–64.

[14] Coventry PA, Hays R, Dickens C, Bundy C, Garrett C, Cherrington A, et al. Talkingabout depression: a qualitative study of barriers to managing depression in peoplewith long term conditions in primary care. BMC Fam Pract 2011;12:10.

[15] Katzelnick DJ, Simon GE, Pearson SD, Manning WG, Helstad CP, Henk HJ, et al.Randomized trial of a depression management program in high utilizers ofmedical care. Arch Fam Med 2000;9:345–51.

[16] Gilbody S, Sheldon T, Wessely S. Should we screen for depression? BMJ 2006;332:1027–30.

[17] Katon W, Unutzer J, Wells K, Jones L. Collaborative depression care: history, evolu-tion and ways to enhance dissemination and sustainability. Gen Hosp Psychiatry2010;32:456–64.

[18] NICE. Depression in adults with a chronic physical health problem: treatment andmanagement. London: National Institute for Health and Clinical Excellence; 2009.

[19] Cannon DS, Tiffany ST, Coon H, Scholand MB, McMahon WM, Leppert MF. ThePHQ-9 as a brief assessment of lifetime major depression. Psychol Assess2007;19:247–51.