A psychometric comparison of military and civilian medical practices

Our purpose was to compare the psychometric properties of military and civilian ambulatory internal medicine care.

Military data came from two cohorts of patients presenting for primary care. Variables collected included mental disorders (the PRIME-MD study), previsit symptom-related concerns and expectations, functional status (Medical Outcomes Study SF-6 and SF-20 scales), postvisit unmet expectations, satisfaction with care (Medical Outcomes Study five-item survey), and physician-perceived "difficulty" (Difficult Doctor-Patient Relationship Questionnaire). This data set was compared with data abstracted from several civilian studies of ambulatory primary care. These studies used the same instruments and took place during the same period.

Military and civilian patient populations were equally likely to have mental disorders, with no differences in the prevalence of disorders within the broad categories of mood, anxiety, somatoform, eating, or alcohol disorders. Civilian populations had a slightly higher rate of some specific diagnoses, including major depression, panic disorder, and generalized anxiety disorder. There was a similar distribution of previsit expectations of care and in the types of postvisit unmet patient expectations, with military patients having slightly lower rates of unmet expectations. Patients in both practice settings had similar self-reported ratings in the six domains of functioning (role, social, pain, emotion, physical, general health), with civilian patients reporting slightly worse overall self-rated health. Both settings had high rates of fully satisfied patients (40% "excellent" for both), with patients with unmet expectations in both groups much less likely to be fully satisfied. There was also a similar proportion of encounters rated as difficult by the clinician (military, 11%; civilian, 15%; p = 0.99). The correlates of difficulty in both groups were similar, with mental disorders and multiple symptoms increasing the likelihood of a difficult encounter.

The psychometric properties of patient care in military and civilian internal medicine ambulatory settings are remarkably similar. These data support mutual generalizability of primary care research findings with respect to psychometric properties between military and civilian populations.