The General Health Questionnaire (GHQ) developed by Goldberg was initially intended to help general practitioners in England to detect cases of nonpsychotic psychiatric illness among their patients. Goldberg himself emphasized that the questionnaire was not designed to detect stable dimensions of personality or character, but nonpsychotic states that might involve relapse, remission or intense transitory changes. The potential uses for this instrument have subsequently grown to include epidemiological studies in the general population, and this has made the GHQ a reference standard among questionnaires used to screen for psychopathological problems in the general population and in nonpsychiatric patients.1
The GHQ is a self-administered instrument with items that deal with subjective symptoms of psychological distress, somatic manifestations frequently associated with anxiety, depression, problems with relationships, and compliance with social, family and professional roles.
Inequalities in health attributable to socioeconomic factors are well known. Clear differences in mortality and morbidity have been found between social classes, and these differences have continued to widen in recent years as social conditions for more disadvantaged populations worsen, especially with regard to economic and employment security. The GHQ also detects these influences in a number of health indicators. For example, the Whitehall II study,2 begun in 1985, involved a cohort of civil servants in London. Participants with a higher job status scored better on health indicators (self-perceived health, long-term illnesses, GHQ, depression, blood pressure, body mass index) than those with a worse job status. These differences were statistically significant in both men and women, with the sole exception of blood pressure in women. The only indicator that was less favorable in higher-status employees was total cholesterol, which was higher in men. The differences in self-perceived health items (SF-36) and the GHQ results were even greater in unemployed persons than in employed persons. Because these comparisons took the degree of economic insecurity into account, the differences in self-perceived health, GHQ score and depression were attenuated in unemployed women. The authors concluded that economic insecurity is even more important than unemployment as a factor that determines health inequalities.
The likelihood that common mental health problems measured by the GHQ will arise, and the chances of recovery, are also determined by social factors. A recent article by Pevalin and Goldberg3 showed that poor social support increases the likelihood of onset of mental health problems and diminishes the chances of recovery. Other social factors such as divorce, losing ones job, remaining unemployed, limitations in daily living activities because of poor health, or caring for a relative also decrease the chances of recovery.
The relationship between psychosocial problems and health service overuse has also been well studied. A report on overuse4 published by members of the Family and Community Medicine Teaching Unit in Jaén (Southern Spain) found significantly more mental disorders among overusers (60%) than among normal users (34%), despite the similarities between the two groups in demographic and family characteristics.
The family physician's ability to identify patients with psychological and social problems in the course of the clinical interview is logically lower than that of a specific instrument such as the Goldberg GHQ. Thus the argument that use of the GHQ increases the number of patients identified with emotional distress is reasonable. In a clinical trial that examined routine inclusion of the GHQ in primary care consultation,5 1589 patients were asked to complete the GHQ in the waiting room, and were then randomly divided into two groups. In the "clinical assessment" group (the physician was not aware of the GHQ results), physicians detected 8.1% of the patients with emotional distress, versus 13.9% who were identified by the group of physicians who had access to the questionnaire results at the start of consultation.
It is interesting to consider which factors aid the physician in detecting psychosocial problems. A Canadian study of the determinants of diagnosis of psychological problems by primary care physicians6 found that psychological problems were detected in 330 (57.7%) of the patients who had a high GHQ score, versus 177 (40.3%) of the patients with a normal GHQ score. In the former group, among patients with a high GHQ score, physicians detected more problems (any sign or symptom of anxiety, depression, somatization or other psychological disorder) if the patient felt that there was a psychological component to his or her problem, and if the physician considered that the reason for consulting was important. In the latter group, i.e., patients who did not have a high GHQ score and whose physician detected a problem, physicians detected more problems (as in the first group) when the reason for consulting included a psychological component according to the patient, or when the physician considered the reason for consulting to be important. In the second group, physicians detected more problems when the patient lived with a partner, the physician knew the patient well, the physician was male, the appointment was scheduled or recommended by the physician, or the problem was familiar from earlier consultations. The authors concluded that for the 50 patients with a normal GHQ score who perceived a psychological component to their problem, detection by the physician seemed to be appropriate. However, in patients with a normal GHQ score who did not perceive any psychological problems, detection by the physician was unlikely to be of any benefit.
The high rate of detection of psychosocial problems (50%) by physicians may lead them to place too much trust in their ability to identify patients with mental disorders. Thus some doctors may be spending unnecessary time and effort on healthy persons, while failing to detect 43% of the patients with high GHQ scores.
The article by de la Revilla et al7 attempts to analyze the usefulness of the GHQ as a method to detect psychosocial problems, in view of the difficulties family physicians face in detecting these patients during primary care visits. These authors show, once again, that psychosocial problems are frequent especially in care settings where patients belong predominantly to the middle-lower class. The relationship between GHQ scores and the results on the Social Readjustment Rating Scale (stressful life events, SLE) shows that recent SLE imply a greater likelihood of high GHQ scores. The GHQ scores were also related with service use. Although the differences after correction for these factors were not significant, the findings nonetheless suggest that problems are more likely in patients such as those studied by de la Revilla et al.
The issue should not be reduced to a question of whether the GHQ helps to detect problems. One of the recommendations of the working group on mental health disorders of the Preventive and Health Promotion Activities Program (PAPPS) for the early detection of anxiety disorders and depression is to use the clinical interview as a basic diagnostic tool. However, at the health professional's discretion, the Goldberg Anxiety and Depression Scale (the GADS, rather than the GHQ) can also be used as a script to help structure the clinical interview8.
As is always the case with early detection measures, it is of fundamental importance to know whether an appropriate response is available for the problems we may detect. Are primary care centers generally equipped to respond to these problems? Most interventions, particularly for anxiety and depression, take the form of pharmacological treatment or brief support psychotherapy. A full report from the Centre for Reviews and Dissemination at the University of York (UK) regarding improvements in the recognition and treatment of depression in primary care9 concluded that the routine administration of questionnaires such as the GHQ improved the diagnosis, but did not improve treatment or the outcome of depression. Their analysis of two systematic reviews concluded that substantial increases would be needed in the role of nurses and in the integration with specialized care for primary care interventions to lead to improvements in the management and outcome of depression.
Key Points
* Many patients with psychological and social problems are identified in primary care. The Goldberg General Health Questionnaire (GHQ) is a useful tool for identifying such patients.
* For the early detection of anxiety disorders and depression, the Preventive and Health Promotion Activities Program (PAPPS) recommends using the clinical interview as the basic diagnostic tool, although at the health professional´s discretion, the Goldberg Anxiety and Depression Scale (GADS) can also be used.
* Routine administration of questionnaires such as the GHQ during primary care consultations improves the diagnosis, but does not improve treatment or the outcome of depression.
* A substantial increase in the role of nurses, and greater integration of primary and specialized care, will be required for interventions to lead to documented improvements in the management and outcome of depression in primary care.
