Several studies have shown that general practitioners do not detect most of their patients’ emotional disorders (Goldberg and Huxley 1980), leading distressed people to consult repeatedly and with multiple professionals. One potential problem is that this phenomenon could increase as a result of free and universal access to healthcare services (Newhouse, Manning, Morris, et al. 1981).

Studies of the relationship between psychological distress and health services utilization have shown that women use healthcare services for affective problems more often than men (Schurman et al. 1985; Fylkesnes, Johnsen, and Forde 1992) and that elderly people obtain fewer services for mental health problems than do their middle-aged counterparts (Leaf, Livingston, Tischler, et al. 1985). On the other hand, inconsistent results are reported regarding the association between income, education, and health services utilization (Fylkesnes, Johnsen, and Forde 1992; Wells, Manning, Duan, et al. 1986). Moreover, studies suggest that individuals who have a usual source of medical care are more likely to use medical health services (Greenley, Mechanic, and Cleary 1987; Leaf, Livingston, Tischler, et al. 1985). Support from family members and friends also influences health services utilization (Horwitz 1978).

In Mechanic’s conceptualization of health services utilization (Mechanic 1978, 1986), individuals perceive a need, evaluate its nature, examine alternatives, and weigh costs and benefits associated with the decision to consult. He also suggests that the perception of this need is influenced by socio-cultural factors. Accordingly, we formulated the hypothesis that a latent construct, “perceived need for health,” constitutes the most immediate cause of illness behavior. This construct reflects the effect of (1) physical and psychological symptoms, (2) predisposing factors encompassing socio-demographic and socio-psychological characteristics, and (3) enabling factors that facilitate or inhibit health services utilization, such as social support (Andersen and Newman 1981). Furthermore, in accordance with the theory that ambulatory medical services utilization is not totally explained by need factors, we hypothesized that predisposing and enabling factors have a direct effect on utilization. Moreover, to statistically identify this hypothesized model, a self-reported perceived health status measure was used to scale the latent construct of perceived need for health. This model could be described in the LISREL notation using the following structural equations:

[x.sub.i] = [[Lambda].sub.i,.sub.m][[Xi].sub.m] + [[Delta].sub.i,sub.i] (1)

[y.sub.j] = [[Lambda.sub.j],.sub.n][[Eta].sub.n] + [[Epsilon].sub.j,j] (2)

[[Eta].sub.n] = [[Beta].sub.n,n][[Eta].sub.n] + [[Zeta].sub.n,] (3)

[[Eta].sub.n] = [[Gamma].sub.n,m][[Xi].sub.m] + [[Zeta].sub.n,n] (4)

These equations are represented in Figure 1, where the [Xi.sub.m] (ksis) are latent constructs representing observed enabling, predisposing and illness variables ([x.sub.i]). The [[Eta].sub.n] (etas) are instrumental latent constructs representing observed outcome variables (y.sub.j). The [[Lambda].sub.i,m] (lambdas) and [[Delta].sub.i,i] (theta-deltas) refer to reliability coefficients and measurement errors of the observed exogenous variables [x.sub.i]. The [[Lambda].sub.j,n] (lambdas) and [[Epsilon].sub.j,j] (theta-epsilons) refer to reliability coefficients and measurement errors of the observed endogenous variables (y.sub.j). The [[Beta].sub.n,n] (betas) represent path coefficients among outcome latent constructs, and the [[Gamma].sub.n,m] (gammas) represent path coefficients among outcome latent constructs and enabling, predisposing and illness latent factors. Finally, the [[Xi].sub.n,n] (zetas) represent variance in the latent constructs explained by external causes not measured in the study.

In the present study we were interested in testing the plausibility of this model in explaining the number of visits to a physician, and to assess the effect of psychological distress symptoms on perceived health need and the use of ambulatory medical services.