[Development of guidelines for rehabilitation of patients with stroke: analysis of therapeutic procedures].

The Federal Insurance Institute for Salaried Employees (BfA, Bundesversicherungsanstalt für Angestellte) in Berlin took the initiative to develop evidence-based guidelines for rehabilitation including stroke rehabilitation. In the present study we performed a systematic survey of the scientific literature on stroke rehabilitation and deduced 8 evidence-based therapeutic modules. They were supplemented by 5 practice-based modules to cover the full range of interventions currently applied in stroke rehabilitation. Modules are clusters of various interventions which were aggregated according to the rehabilitative goal they have in common. Data were analysed from 8,876 BfA patients, the total population of "neuro"-patients in 2001 and 2002 with 142,951 interventions and 1,071,885 appointments. Stroke diagnoses covered ICD-10 codes I60, I61, I62, I63, I64 or were coded by G45, G81 and one of these I-codes. Stroke rehabilitation as provided in daily practice was found to be a complex multimodal treatment programme in all hospitals which were studied. On average, patients receive treatments from 7.3 therapeutic modules. A dominating role is played by "motor modules". Obviously, they play a major role in acute stroke rehabilitation in all hospitals compared to cognitive retraining, language therapy and especially improvement of psychic functions, which were significantly less applied. A dramatic variation could be shown to exist between the rehabilitation hospitals (1) with respect to the number of patients who were treated with a given module (e. g., some modules showed variations between 18 % and 100 %), and (2) with respect to treatment intensity and treatment duration per week. Several factors can be taken into account for interpretation of this high variance. Intrinsically, the high variation in symptomatology after stroke can play a major role, as both the deficit profile and the level of severity of a given deficit and the combined severity level can vary from patient to patient. On the other hand, extrinsic factors such as lack of staff may also account for some of the differences. In any case, the results show the necessity to optimize stroke rehabilitation by development and implementation of guidelines. Yet, they also point to a need for further research into rehabilitation interventions and procedures to yield evidence for the approximately 30 % of "only" practice-based interventions.