This proposed 3 year research study seeks to determine the differences in patient outcomes between patients who contract to practice behavioral strategies in order to manage their chronic illness and those patients who write contracts with the nurse in order to improve compliance with specific, prescribed therapeutic regimens. The questions to be addressed in this proposal are to determine whether improved patient outcomes can be enhanced most by: 1. writing contracts for behavioral strategies after the patient has received instructions about such strategies or by, 2. writing contracts for behavioral strategies without prior instructions or by, 3. writing contracts to increase compliance behaviors to specific, prescribed, therapeutic regimens. Four hundred patients with diabetes will be assigned to one of four groups. These groups will consist of 1. No Contract Group, 2. Contracting and Instructions for Behavioral Strategies, 3 Contracting for Behavioral Strategies, 4. Contracting for Compliance. Patients will be chosen over a two year period. During these first two years each year will be divided into four blocks of three months each. Random choice will determine which of the four patient groups will be studied in a given block. This design will minimize history and ordering effects since the nature of the diabetic unit is such that only one patient group can be studied at a time. The contracts will be written by five nurses currently employed on the unit. These nurses will be randomly assigned to the patients. Outcome measures will include capillary glucose monitoring. Glycosolated hemoglobin (HbA1), the Goldenrod Scale developed by Kaiser Permente, the Goldenrod Scale of Perceived Health, weight, patient knowledge and the number of clinic appointments kept. Profiles will be developed relative to the contracts. The number and sequence of contracts will be tested as a dependent measure. The study will provide important information about specific mechanisms for improving patient outcomes.
