Impact of a critical pathway on the results of carotid endarterectomy in a tertiary care university hospital: Effect of methods on outcome

Purpose: A carotid endarterectomy critical pathway (CP) targeting a 3- day postoperative course was introduced in March 1994. This retrospective analysis assesses its impact on operative results, postoperative length of stay (POD), and cost of hospitalization (COH). Methods: One hundred eighty- six patients who underwent 201 carotid endarterectomy procedures from Nov. 1992 to Feb. 1994 (Pre-CP; n = 67)and from Apr. 1994 to Jul. 1995 (Post-CP; n = 134) at Johns Hopkins Hospital, a tertiary care referral center, were evaluated. Results: The Pre-CP and Post-CP groups had similar risk factors, postoperative morbidity rates, and mortality rates. Furthermore, they had similar mean POD (Pre-CP, 6.0 ± 0.5 days; Post-CP, 5.7 ± 0.6 days; p = 0.79) and COH. However, only 85 of the Post-CP (63%) patients were actually placed on the CP (CP-starters); the mean POD was 3.4 ± 0.3 days among these CP-starters (p < 0.0001) and 2.8 ± 0.1 days among the 74 Post-CP patients (55%) that remained on the pathway (CP-finishers; p < 0.0001). The mean COH was reduced from $12,881 (Pre-CP) to $9701 for the CP-starters (p = 0.01) and to $8572 for the CP-finishers (p = 0.0001). However, we found that only 47 of the Pre-CP patients (70%) would have been eligible for the CP, and the mean POD among those cases was 4.2 ± 0.4 days, which was not different than the mean POD among the CP-starters (p = 0.17). The mean COH of the eligible Pre- CP cases, $9508, was not significantly different from the COH of the CP- starters (p = 0.97). Conclusions: This subset analysis emphasizes the importance of establishing an accurate 'control' group when studying a CP, because using all of the Pre-CP cases as the 'control' group in the original analysis, including patients who would not have been candidates for the CP, clearly overstated the beneficial impact of the CP.