Abstract
Ariel Eytan and Christel Alberque, Department of Psychiatry, Geneva University Hospitals, Geneva, Switzerland:
Being in charge of a general psychiatry unit receiving hospitalized patients with somatic comorbidity, we feel that Dr Viinamäki's paper in the October 2000 issue of this Journal [1] raises interesting methodological questions.
Contrary to a previous study showing a lower recovery rate of patients with compound depression than for patients with pure depression [2], Viinamäki fails to demonstrate a significant effect of somatic comorbidity on recovery from depression in a 6-month follow up. We are concerned with the definition of somatic comorbidity in this context and believe it might account for this negative result. Indeed, his definition is so broad (‘the patient had at least one diagnosed and, in the short term, non-fatal physical disease’) that it encompasses a wide range of conditions with very different degrees of severity. As a consequence, the impact of the physical illness in terms of level of functioning, impairment and anxiety varies a great deal from one patient to another. Moreover, a majority of patients in Dr Viinamäki's study suffered from musculoskeletal diseases, mainly spondyloarthrosis. This disorder is a well-known cause of chronic pain. As such, it should be differentiated from other nonpainful chronic disorders because pain interacts with depression in various ways: Depression may result from poor outcome of a pain disorder. Conversely, depression is sometimes overlooked and causes refractoriness to pain treatment in chronic physical diseases [3].
Specifying that both problems (somatic and psychiatric) must require active investigations and treatment at a given moment is a way to refine the above definition of comorbidity in a pragmatic way. This precision has indeed direct implications on the organization of care, especially in case of hospitalization: it is advisable to refer a patient that fulfills this criterion to a psychiatric service closely connected with a general hospital in order to optimize a global treatment plan. By emphasizing dynamically the overlapping of morbid conditions, the focus is on episodes of exacerbation of chronic diseases and on nonlife-threatening acute disorders rather than on chronic and asymptomatic physical conditions (like, for example, arterial hypertension well-controlled by a current medication). Nevertheless, this gain in relevance is not sufficient and the severity of the physical illness remains to be assessed. In order to characterize this dimension, we recommend the utilization of a standardized instrument like the MOS 36-Item Short-Form Health Survey (SF-36) [4]. This scale assesses eight health concepts: (i) limitations in physical activities because of health problems; (ii) limitations in social activities because of physical or emotional problems; (iii) limitations in usual role activities because of physical health problems; (iv) bodily pain; (v) general mental health (psychological distress and wellbeing); (vi) limitations in usual role activities because of emotional problems; (vii) vitality (energy and fatigue); and (viii) general health perceptions.
Literature suggests that physical illness is associated with increased length of stay (LOS) for depressed psychiatric patients but not for other specific diagnostic groups [5]. Among the first 292 patients hospitalized in our unit, 197 of them had a somatic comorbidity using the definition we propose. Their LOS was longer than for the purely psychiatric patients, but the difference is of borderline statistical significance (mean = 19.5 days vs mean = 15.2 days). A subgroup 147 patients suffered from depression according to ICD-10 guidelines. Among them, 95 had a somatic comorbidity. Their LOS was longer than for the purely psychiatric patients (mean = 21.3 days vs mean = 17.7 days). These preliminary data tend to indicate that, in our population, LOS is increased by the presence of physical illness, whatever the psychiatric diagnosis.
Coexistence of psychiatric and somatic illnesses is an important issue: it is frequent and very probably impacts on LOS and clinical outcome. The scarcity of empirical data stresses the necessity to characterize more accurately the somatic component of illness in psychiatric studies.