Assessing the Validity of the Brief Pain Inventory for Use Among Adults with Sickle Cell Disease

Pain is a primary concern for adults living with sickle cell disease (SCD). Accurately assessing pain severity and the extent to which it interferes with physical and social functioning in daily life is an important first step in treating pain and appropriately addressing patients' needs. Several multidimensional self-report pain scales exist, however, the Brief Pain Inventory short-form (BPI) has become one of the most widely used measurement tools for assessing pain in the clinical setting. The instrument assesses both pain intensity and interference; it has been proven to be a reliable and valid measure for assessing pain in diverse settings, has been translated into numerous languages, and has been utilized in different ethnic and cultural groups. As a self-report scale, the BPI is both low cost and relatively easy to administer and score. However, the BPI, like all self-report scales, is dependent on accurate patient comprehension of all items for the scale to be unbiased and valid. To our knowledge, no studies have tested the validity of the BPI among adult patients with SCD, a group that may present with a diversity of sociocultural and literacy levels affecting how they complete self-report questionnaires. Therefore, the objective of this study was to determine the validity and reliability of BPI scores in patients living with SCD.

Methods: The brief pain inventory was administered to adults receiving routine care at a comprehensive sickle cell clinic. All completed BPI forms were examined and independently scored by two trained reviewers who also checked for the following types of input errors: 1) Patients indicated experiencing no pain in the past week on a yes-or-no item but subsequently indicated pain location and/or pain severity greater than zero on a numeric scale; 2) Data entered for numeric scales of worst, least, average and current pain conflicted with each other; 3) Patients did not appropriately complete the body image by shading all pain areas and marking an 'X' on the body image item for the worst pain area; 4) Patients had at least one missing item. In addition, reviewers identified patients who responded in a non-traditional way by indicating scores that lie between numbers on the scale, selecting multiple numbers on the scale, or indicating a pain level above ten on the scale. Disagreements on categorization of errors were resolved by a third reviewer.

Results: Sixty-nine adults completed the BPI, 54% were female, and the average age was 36 (range 23-60) years. Average pain intensity across all items was 4.7, average interference score was 4.6. As many as 88% of the pooled BPI assessments contained some form of error or combination of errors: 22% of patients indicated no pain but reported above-zero pain intensity or above-zero pain interference, 51% had conflicting scores on the pain intensity numeric scales, 80% had completion errors on the body image item, and 45% had one or more missing items. Twelve patients responded non-traditionally (e.g., indicating >10 on a 0 to 10 scale item.)

Discussion: This systematic examination of completed BPIs among adults with SCD found that a majority of patients are not completing this well-validated, short, self-report pain scale correctly. This indicates that studies or clinical programs using the BPI to assess pain in SCD may be receiving invalid results.

These errors likely stem from some combination of insufficient patient comprehension level, low literacy, sociocultural differences, or ambiguity of the BPI items' language. Whatever the reason for these errors, the high frequency of these documented issues in this sample of sickle cell patients is concerning, considering the widespread use and published reliability and validity of this scale.

Limitations: It should be noted that this is a study of a relatively small sample from one geographic location (i.e., Pittsburgh, PA). Such errors with the BPI may or may not arise in other SCD populations.

Despite its limited scope, our study suggests that more careful consideration is warranted for use of the BPI in SCD and other typically low-income, or low health literacy populations, which may pose high risk for inaccurate completion of the measure and potential bias. Suggestions for revisions to the BPI include simplifying the language, shortening sentence length, and more clearly specifying the time-frame for each item.