As part of a study on treatment of uncomplicated cellulitis, outpatients age 12 seeking care at the emergency department with cellulitis and no wound, purulent drainage, or abscess were randomly assigned to either cephalexin plus trimethoprim-sulfamethoxazole for 7 days (n = 85) or cephalexin plus placebo for 7 days (n = 87). The primary outcome was clinical cure, defined as absence of fever and more than minimal erythema, swelling, or tenderness (days 14-21). A total of 24 patients in the cephalexin plus trimethoprim-sulfamethoxazole group and 18 patients in the cephalexin plus placebo group did not complete treatment as prescribed or failed to complete the test-of-clinical-cure visit; they were categorized as noncompliant patients. The principal investigator of the study wanted to conduct a per-protocol analysis of the data. Which of the following describes how the data pertaining to noncompliance should be treated to conduct such an analysis?

Question

As part of a study on treatment of uncomplicated cellulitis, outpatients age >12 seeking care at the emergency department with cellulitis and no wound, purulent drainage, or abscess were randomly assigned to either cephalexin plus trimethoprim-sulfamethoxazole for 7 days (n = 85) or cephalexin plus placebo for 7 days (n = 87).  The primary outcome was clinical cure, defined as absence of fever and more than minimal erythema, swelling, or tenderness (days 14-21).  A total of 24 patients in the cephalexin plus trimethoprim-sulfamethoxazole group and 18 patients in the cephalexin plus placebo group did not complete treatment as prescribed or failed to complete the test-of-clinical-cure visit; they were categorized as noncompliant patients.  The principal investigator of the study wanted to conduct a per-protocol analysis of the data.  Which of the following describes how the data pertaining to noncompliance should be treated to conduct such an analysis?

Quizplus · Accepted Answer

Explanation: 11ec2b69_bdfc_bbfa_8f63_bf6e4ded6857_MD0005_00 Per-protocol (PP) analysis compares treatment groups in a randomized trial by including in the analysis only those participants who strictly adhered (eg, compliant participants) and completed the protocol. Noncompliant participants may bias estimates of treatment effect. For example, in a placebo-controlled randomized trial of a drug (ie, treatment vs control group), adverse effects may cause a greater proportion of sicker patients to drop out from the treatment group than from the control group. At the trial's conclusion, the treatment group may end up with a smaller proportion of more severely ill patients than the control group (eg, selection bias), and this result could distort the estimation of the effect of a treatment (eg, confounding bias). By excluding all noncompliant participants, PP analysis provides an estimate of the true effect of an intervention assuming a perfect scenario in which all participants were retained and completed treatments as planned. However, this overestimates the effect of the intervention in a realistic (ie, practical) clinical setting, which almost inevitably includes noncompliant patients. In accordance with PP analysis, all 18 + 24 = 42 noncompliant participants should be excluded from analysis (Choices A, B, and C). In contrast, an analysis that includes all participants as initially allocated after randomization, regardless of what happens during the study period (ie, all 42 noncompliant participants in their respective groups), follows an intention-to-treat (ITT) principle (Choice E). ITT analysis could lead to a more conservative estimate of the effect of the intervention; however, because it closely mirrors the expected number of noncompliant patients in a real-life situation, ITT analysis reflects the expected effect of the intervention in a realistic (ie, practical) clinical setting. Educational objective: Per-protocol analysis includes only those participants who strictly adhered (ie, nondropouts) to the study protocol. This approach tends to provide an estimate of the true effect of an intervention for a perfect scenario, but it overestimates the effect of the intervention in a practical clinical setting.

As Part of a Study on Treatment of Uncomplicated Cellulitis