Outcome measurement in clinical practice: practical and theoretical issues for health related quality of life (HRQOL) questionnaires

Abstract

The defining characteristic of clinical outcome measurement tools is that they are intended to measure change over time. There is a range of practical issues that must be considered when choosing outcome measures for routine use in the clinical setting. Outcome measures should be: relevant to patients' values and perceived needs; measure aspects of health and function that are relevant to the treatment offered; have sound psychometric properties; be quick and easy to complete and score; and, be intuitive to interpret. They should also provide clinically meaningful information useful to the clinician for decisionmaking regarding patient care. Outcome measures fall broadly into two types: measures of impairment of physical structures or functions; and measures of activities or participation as determined by health-related quality of life (HRQOL) questionnaires. Whilst recognizing that measures of physical impairments have an important role in patient assessment, this paper focuses on HRQOL measures. It is intended as a summary of the key attributes of HRQOL outcome measurement tools. The different types of outcome measures are reviewed along with the important scientific and practical considerations for the choice of an outcome measure. The determination of validity, reliability, responsiveness for outcome measures is explained, and the interpretation of changes in score for an outcome measure in terms of minimal important difference is outlined.

Reliability and Validity of the Dutch Version of the International Physical Activity Questionnaire in Patients After Total Hip Arthroplasty or Total Knee Arthroplasty

“…The SEM reflects the amount of error associated with an individual patient score. 19 MDC 95 is an estimate of the lowest change value that can be detected beyond the random measurement error. 19 To calculate the SEM, the standard deviation of the first administration (population at baseline) was multiplied by ͌(1 -r) , where r represented the ICC.…”

“…19 MDC 95 is an estimate of the lowest change value that can be detected beyond the random measurement error. 19 To calculate the SEM, the standard deviation of the first administration (population at baseline) was multiplied by ͌(1 -r) , where r represented the ICC. 8 To calculate the MDC 95 at the population level, the fol-lowing formula was used: (1.96 × 2 × SEM)/N.…”

Minimum Important Differences for the Patient-Specific Functional Scale, 4 Region-Specific Outcome Measures, and the Numeric Pain Rating Scale

“…Patient-reported outcome measures are ubiquitous, but interpretation of the meaningfulness of change between one administration and another requires an understanding of the measurement properties of the outcome measure. 12 The most important clinimetric property for interpreting change over time is the minimum important difference (MID), also called the minimum clinically important difference or minimum important change. 12 The MID is best established against a patient-rated external "anchor," but may vary depending on the external anchor that defines patient-perceived importance.…”

“…12 The most important clinimetric property for interpreting change over time is the minimum important difference (MID), also called the minimum clinically important difference or minimum important change. 12 The MID is best established against a patient-rated external "anchor," but may vary depending on the external anchor that defines patient-perceived importance. Currently, there is no consensus regarding the cut point of the external anchor that best defines important change.…”

Validity of pain drawings for predicting psychological status outcome in patients with recurrent or chronic low back pain

“…21 To independently verify these previously described cut-off values, we calculated the optimal cut-off value for PSS using receiver operating characteristics (ROC) curve methodology, described elsewhere. 31,32 The above analyses were conducted using the SPSS computer program (SPSS Inc., Chicago, IL, USA). Contingency tables were used to calculate RR.…”