Does maintenance lithium therapy prevent recurrences of mania under ordinary clinical conditions?

Mental-hospital admission rates in Edinburgh for mania, schizophrenia and psychotic depression were studied from 1970 to 1981, a 12-year period during which long-term lithium therapy was increasingly employed in affective illnesses. If this treatment had been effective admission and readmission rates for mania, and perhaps also for depression, should have fallen progressively. In fact, they rose steadily, while the admission rate for schizophrenia fell. These changes could not easily be attributed to changing diagnostic criteria, to the admission of milder affective illnesses, or to poor and deteriorating lithium surveillance. Their explanation is uncertain, but they cast some doubt on the efficacy of lithium prophylaxis in ordinary clinical practice.