Per-protocol analysis

The per-protocol analysis (analogous to "analysis according to protocol, evaluation according to test plan ") is an evaluation method of medical statistics that is mainly used in the context of randomized controlled intervention studies . The shortened term per protocol or the shortened English terms per protocol or as per protocol are often found in medical literature .

method

In the per-protocol analysis, only the results of those subjects who behaved in accordance with the test plan are taken into account when evaluating the data , i.e. H. which according to protocol (ie per-protocol (strictly speaking english per protocol )) assigned to treatment with the active treatment or the placebo also actually received. Results from subjects who - for whatever reason - dropped out of the study prematurely, discontinued the intervention (medication, procedure, etc.), switched to another study group, or otherwise failed to fully comply with the study protocol , are therefore not included in the analysis. The advantage of this evaluation method is that it can be used to estimate the effectiveness potential of an intervention under optimal conditions. One disadvantage, however, is that the randomization principle is broken, so that a statistical distortion of the results (bias) can occur, which then often turns out to be in favor of the intervention.

Because of its disadvantages, the per-protocol analysis is mainly used in intervention studies of phases I to II , in which first of all tolerability, pharmacological properties, therapeutic effectiveness and optimal dosage are to be determined.

The intention-to-treat analysis should be mentioned as a supplement or alternative to a per-protocol analysis . Here, the data of all test subjects are evaluated, regardless of whether they have behaved in accordance with the test plan or not.

The as-treated analysis is a subtype of the per-protocol analysis. The data of all those test persons are evaluated who received the intervention or medication to be tested in the context of the study, regardless of which study group they were originally assigned to.

Web links

Individual evidence

↑ Hans-Heinrich Raspe et al .: Recommendations for the review of clinical studies by ethics committees. 2nd, revised and updated edition. Deutscher Ärzte-Veralg Cologne, 2012, ISBN 978-3-7691-1305-1 , p. 184.

[1] Hans-Heinrich Raspe et al .: Recommendations for the review of clinical studies by ethics committees. 2nd, revised and updated edition. Deutscher Ärzte-Veralg Cologne, 2012, ISBN 978-3-7691-1305-1 , p. 184.