COMPREHENSION AND READABILITY OF DRUG INFORMATION: A COMPARATIVE STUDY AT DIFFERENT LEVELS OF READING ABILITY.

Abstract

Ley's Partial Model of Compliance suggests that patients who understand information given to them are more likely to remember the information and are more likely to be satisfied with the information. The model then suggests that these components will lead to greater patient compliance with medication regimens. To test the model, Patient Package Inserts (PPIs) describing thiazide diuretics from the American Association of Retired Persons, the American Medical Association, the Canadian Pharmaceutical Association, the Food and Drug Administration, the National Association of Retail Druggists, the United States Pharmacopoeial Convention and a Test PPI written by the Principal Investigator were used. The SMOG Readability Formula was used to determine the grade levels at which PPIs were written. One hundred thirty-six adults enrolled in GED classes in Tucson and other communities and 107 adults enrolled in remedial reading classes at Tucson's Pima Community College were administered the Zip Scale reading placement test and blocked by their reading abilities. Within each of the three blocks, subjects randomly received one of the seven information sheets or no sheet. Subjects took a multiple-choice test based upon information common to all of the PPIs, a cloze comprehension test based upon the PPI which they read, and completed a satisfaction survey which asked subjects to rate the PPI which they read. Subjects also read five vignettes describing fictitious patients taking thiazides who were confronted with different barriers to compliance. Subjects indicated how likely the fictitious patients were to overcome the barriers to compliance. Among this sample of remedial-reading adults, the Test PPI emerged as clearly superior to the others for any of the variables measured. This result would behoove providers of PPIs to rewrite PPIs, reducing the difficulty of these documents as much as possible. Ley's Partial Model of Compliance did not accurately describe the associations between Understanding, Memory, Satisfaction and Compliance for this sample. A New Model emerged describing different associations between these components and between subject reading ability and PPI readability.