Quantitative measures of rheumatic diseases for clinical research versus standard clinical care: differences, advantages and limitations

No single measure can serve as a ‘gold standard’ for the diagnosis, prognosis, and monitoring of patients with rheumatic diseases. Therefore, pooled indices of several measures have been developed for patient assessment. Quantitative measures and indices in rheumatology have been used primarily in clinical trials and other clinical research, but not in standard clinical care. Indeed, most standard rheumatology care is conducted without quantitative data other than laboratory tests, which often are uninformative. Some measures used in research have been adapted for standard care. The classical 66/68-joint count with graded scoring for swelling, tenderness, pain on motion, limited motion, and deformity has been shortened for clinical care to a 28-joint count, scored only as ‘Yes’ or ‘No’ for swelling or tenderness. Patient questionnaires designed for clinical research can be lengthy, with complex scoring, so that information is not available to help guide clinical decisions. By contrast, patient questionnaires designed for standard care, such as a simple one-page, multi-dimensional health assessment questionnaire (MDHAQ), are short, save time, are easily scored, and are useful in all rheumatic diseases to monitor patient status at each visit and document changes over long periods. More attention to measures for use in standard care could improve care and outcomes for patients with rheumatic diseases.

Introduction

Quantitative clinical measures ranging from temperature to serum glucose have greatly advanced clinical care of most diseases. Translation of qualitative clinical impressions into quantitative data contributes to improvements in diagnosis, prognosis, and treatment. Quantitative assessment in rheumatic diseases differs from many clinical conditions in that a single ‘gold standard’ measure such as blood pressure or serum cholesterol, which can be used to assess all individual patients in clinical trials, clinical research, and clinical care, is not available. Therefore, pooled indices of multiple measures¹ have been developed, such as the American College of Rheumatology (ACR) Core Data Set2, 3, 4 and Disease Activity Score (DAS)5, 6 to assess and monitor patients with rheumatoid arthritis (RA). Indices have also been developed to assess systemic lupus erythematosus (SLE)7, 8, 9, 10, 11, 12, 13, 14, vasculitis15, 16, 17, 18, 19, 20, psoriatic arthritis21, 22, 23, ankylosing spondylitis24, 25, 26, 27, 28, and other diseases (Table 1).

Considerable advances in quantitative measurement in rheumatology over the last two decades have been applied primarily to clinical trials and other clinical research. One matter that has not received substantial attention is the application of measures in standard rheumatology clinical care. Indeed, in contrast to clinical trials, most standard rheumatology care continues to be conducted largely according to qualitative ‘Gestalt’ impressions, without quantitative data other than laboratory tests, which frequently give false-positive, false-negative, or non-informative results.29, 30

Management of inflammatory rheumatic diseases without quantitative information may be regarded as analogous to giving treatment for a fever without a temperature, rapid heart rate without a pulse, or even elevated blood pressure or serum glucose without quantitative information. Such clinical management can be effective, but quantitative data enhance effective care, as well as documentation of results.

Measures and indices designed for clinical trials and other clinical research can differ substantially from measures designed for standard clinical care (Table 2). The primary criteria for acceptability of measures in clinical research are validity – ‘Does the measure address what is thought to be measured?’ – and reliability – ‘Is the measure reproducible?’31, 32 Measures to be used in standard clinical care must be valid and reliable, but also feasible and acceptable to patients and health professionals. For example, the classical 66/68-joint count, with five graded criteria for swelling, tenderness, pain on motion, deformity, and limited motion, has been abbreviated to 28 joints that are scored ‘Yes’ or ‘No’ for swelling and tenderness. Elaborate patient questionnaires that were not designed to be reviewed and scored by a clinician when treating patients [even including the Health Assessment Questionnaire (HAQ)³³] have been modified to instruments such as the Multidimensional HAQ (MDHAQ)34, *35, which is easily scored to be available for clinical decisions in standard care. Furthermore, an index based on the MDHAQ, termed routine assessment of patient index data 3 (RAPID3), which can be scored in 10 seconds, depicts status and differences between active and control treatments in clinical trials as effectively as a DAS.36, 37, *38

This chapter describes measures used to assess patients with RA, including joint counts, radiographs, laboratory tests, patient questionnaires, and global measures, as well as RA indices. Measures and indices used in research settings (including clinical trials), in which the primary criteria are validity and reliability, are contrasted with measures for standard clinical care, which require the additional considerations of clinical utility, feasibility, and acceptability. Some advantages and limitations of each of the different types of measures are discussed (Table 3). Further details are found in previous articles that review measures used in rheumatic diseases.39, 40, *41, *42, 43