Unclear neurological illness in neurological inpatients

Objectives: To compare the clinical profile of hospitalized patients with neurological complaints who remained undiagnosed upon discharge with those who are given a firm neurological diagnosis. Methods: A retrospective, case-control study on patients matched by age, sex and year of admission was carried out in a 1172-bed teaching hospital. Group A (n = 80) included patients hospitalized between 1991 and 1993 with neurological complaints considered by neurologists of 'non-neurological origin' or diagnosed of a non-neurological disease. Group B (n = 80) included matched patients admitted to the Neurology ward during the same period who received a firm neurological diagnosis. Results: Patients in Group A had suffered from more stressful events during the past 3 months, had had more psychiatric referrals and/or recommendations for psychiatric follow-up. Patients in Group B were more investigated, and more of their results were abnormal; they also stayed longer in hospital, received more medication and were seen more often in the neurological follow-up. There was no difference between the two Groups in regards to pain and digestive tract complaints. Memory complaints of 'non- organic' type and somatic anxiety were significantly more common in Group A. Exploratory Factor Analysis of data from Group A yielded four factors accounting for 43% of variance which in terms of their predominant loadings were called: anxiety, migraine, somatoform and pseudo-epileptic. Conclusions: Hospitalized patients with neurological complaints but unclear neurological diagnosis are distinguishable from matched counterparts with firm neurological diagnosis. They cluster into sub-types that make sense clinically. Prospective studies are needed to ascertain our findings.