The efficacy of 0.9% sodium chloride injection with and without heparin in maintaining indwelling intermittent ("heparin lock") injection sites was studied. All patients in whom heparin locks were placed after admission to the medical and surgical units of a 128-bed acute-care hospital during a six-month period were included in the study. Three different solutions were used to flush heparin locks: 0.9% sodium chloride injection alone, heparin 10 units/ml in 0.9% sodium chloride injection, and heparin 100 units/ml in 0.9% sodium chloride injection. Solutions were randomly assigned to all patients on a given nursing unit for a two-month period; flush solutions were switched every two months until each of the three solutions had been used on both the medical and surgical units. Heparin locks were flushed after each medication administration and every eight hours when medications were not being given. Using a standardized evaluation form, one of five i.v. therapists evaluated each site daily for the presence of phlebitis and loss of patency. Length of catheter placement and the percentage of patient days during which patients received cephalosporin and penicillin antibiotics were examined for each group. Rates of site loss caused by phlebitis or loss of patency were compared in each group. A total of 412 patients representing 1448 patient days of heparin-lock therapy was evaluated. No significant differences were found among the three groups in the mean duration of heparin-lock placement, the percentage of patient days during which antibiotics were prescribed, or the rate of site loss caused by phlebitis or loss of patency.(ABSTRACT TRUNCATED AT 250 WORDS)