Abstract. Ragnarson Tennvall G, Apelqvist J, Eneroth M (Swedish Institute for Health Economics and Departments of Internal Medicine and Orthopaedics, Lund University Hospital, Lund, Sweden). The inpatient care of patients with diabetes mellitus and foot ulcers. A validation study of the correspondence between medical records and the Swedish Inpatient Registry with the consequences for cost estimations. J Intern Med 2000; 248: 397–405.
Objectives. To compare the correspondence of discharges and diagnoses between medical records and an administrative database in diabetic patients with foot ulcers and to calculate inpatient costs from the different sources.
Design. Discharge data from 117 prospectively followed patients were compared with information from the Swedish Inpatient Registry during the same treatment period for each patient. Non corresponding discharges were retrospectively reviewed in patient records. Costs of inpatient care were calculated from clinical sources and from data selected in three different ways from the database. Information in medical records was assumed to represent the ‘Gold Standard’.
Setting. In-hospital treatment, multidisciplinary foot-care team.
Subjects. A total of 117 diabetic patients with deep foot infections.
Main outcome measures. Degree of database completeness and inpatient costs.
Results. The degree of completeness of discharges in the database was 98%, but 8.6% of discharges had no code for diabetes and 13% were registered without any foot related diagnosis code. Less than 20% of discharges were found with selection based on primary diagnosis 250.G only. Total inpatient costs varied from 2.7 to 13.3 million SEK (Swedish Kronor) depending on selected diagnosis codes. Approximately 84% of all foot ulcer discharges could be found in the database if codes with both diabetes mellitus and foot ulcer diagnoses were selected.
Conclusions. Inpatient costs for foot complications cannot be accurately estimated from the Inpatient Registry when based on primary diagnosis exclusively. Fairly good estimates at a low data acquisition cost can be made with a combination of foot related diagnoses together with codes for diabetes.