Comparison of a planned versus emergent approach implementation of tight glycemic control therapy on an ICU: an experimental design

Introduction

In this study the effectiveness of a hierarchic planned approach to implementation was compared with an emergent approach with a high degree of nurse participation. The innovation was a new tight glycemic control (TGC) therapy in the ICU. The implementation method to introduce a new evidence-based treatment to the ICU can make a difference to its effectiveness. New treatment procedures are very often strictly defined and introduced on the ICU in a top-down manner. On theoretical grounds, a bottom-up implementation method, with more participation and involvement of the ICU nurses, should be more effective.

Methods

In an experiment a TGC protocol was introduced on a 24-bed ICU with 120 ICU nurses. In one nursing team the implementation was by means of the usual planned approach (PA). In the other team an emergent implementation approach (EA) was applied. During 7 months, 1,182 patients were admitted and randomly assigned to either team. The effectiveness of the implementation methods was measured at patient-outcome (glucose) level and on a process-execution (nurse compliance) level. The patient-outcome measurements were: mean glucose value, percentage of normal glucose values (defined as between 4.0 and 6.5 mmol/l) and the time from admission to reach a normal glucose level (6.5 mmol/l). For the mean glucose value a difference was made between short-stay (<3 days) and long-stay patients (≥3 days). The compliance was expressed as compliance for safe, insulin treatment (the patient received the right amount of insulin) and compliance for efficient treatment (glucose measurements as well as insulin were according to the protocol).

Results

The mean glucose for the short-stay patients was significantly lower in the EA than the PA team (6.42 vs 6.52 mmol/l), no significant difference was found for the long-stay patients. The percentage of normal glucose values was 53.9% in the EA and 52.8% in the PA team (significant P < 0.001). The time to reach normal glucose levels was 6 to 7 hours in the EA and 10 to 11 hours in the PA team (significant, P < 0.05). The compliance for safe insulin treatment was 91.3% in the EA and 79.0% in the PA team. The compliance for efficient treatment was 83.5% in the EA and 66.8% in the PA team.

Conclusions

An emergent approach to implementation (with an increased participation of the ICU nurses) of a TGC resulted in better patient results and a higher compliance to the protocol.