Catheter-associated bloodstream infections and urinary tract infections are frequently encountered health care-associated infections. We aimed to reduce inappropriate use of catheters to reduce health care-associated infections.

In this multicentre, interrupted time-series and before and after study, we introduced a de-implementation strategy with multifaceted interventions in seven hospitals in the Netherlands. Adult patients admitted to internal medicine, gastroenterology, geriatic, oncology, or pulmonology wards, and non-surgical acute admission units, and who had a (central or peripheral) venous or urinary catheter were eligible for inclusion. One of the interventions was that nurses in the participating wards attended educational meetings on appropriate catheter use. Data on catheter use were collected every 2 weeks by the primary research physician during the baseline period (7 months) and intervention period (7 months), which were separated by a 5 month transition period. The primary outcomes were percentages of short peripheral intravenous catheters and urinary catheters used inappropriately on the days of data collection. Indications for catheter use were based on international guidelines. This study is registered with Netherlands Trial Register, NL5438.

Between Sept 1, 2016, and April 1, 2018, we screened 6157 patients for inclusion, of whom 5696 were enrolled: 2650 patients in the baseline group, and 3046 in the intervention group. Inappropriate use of peripheral intravenous catheters occurred in 366 (22·0%, 95% CI 20·0 to 24·0) of 1665 patients in the baseline group and in 275 (14·4%, 12·8 to 16·0) of 1912 patients in the intervention group (incidence rate ratio [IRR] 0·65, 95% CI 0·56 to 0·77, p<0·0001). Time-series analyses showed an absolute reduction in inappropriate use of peripheral intravenous catheters from baseline to intervention periods of 6·65% (95% CI 2·47 to 10·82, p=0·011). Inappropriate use of urinary catheters occurred in 105 (32·4%, 95% CI 27·3 to 37·8) of 324 patients in the baseline group compared with 96 (24·1%, 20·0 to 28·6) of 398 patients in the intervention group (IRR 0·74, 95% CI 0·56 to 0·98, p=0·013). Time-series analyses showed an absolute reduction in inappropriate use of urinary catheters of 6·34% (95% CI −12·46 to 25·13, p=0·524).

Our de-implementation strategy reduced inappropriate use of short peripheral intravenous catheters in patients who were not in the intensive care unit. The reduction of inappropriate use of urinary catheters was substantial, yet not statistically significant in time-series analysis due to a small sample size. The strategy appears well suited for broad-scale implementation to reduce health care-associated infections.

Netherlands Organisation for Health Research and Development.