SURGERY J U LY
Surgical research review Using quality-of-life instruments to assess surgical outcomes Vic Velanovich, MD, Detroit, Mich
From the Division of General Surgery, Department of Surgery, Henry Ford Hospital, Detroit, Mich
ASSESSING SURGICAL OUTCOMES HAS BECOME an area of intense interest among surgeons, payors, and administrators. Although there are several aspects of what makes up an “outcome” (and some of these aspects are vague), there is no question that quality of life is one of these aspects. Quality of life has been receiving additional attention in both the medical and surgical literature. Despite this, physicians and patients can differ substantially in the assessment of the patient’s quality of life.1 This is true because the concepts and mechanisms concerning the measurement of quality of life may still be confusing to some. Quality of life is a concept easy to grasp intuitively but difficult to define. In the common sense of the phrase, quality of life may include “nonmedical” factors such as personal income, location of residence, leisure time and activities, interpersonal relationships, meaningful work, and spiritual fulfillment, among other aspects of everyday life. However, in medicine and surgery what we wish to measure is “health-related” quality of life, that is, the extent of the impact of the disease process on physical, psychologic, and social aspects of a person’s life and feeling of well-being. These different aspects of quality of life are called domains. Qualityof-life “instruments” can measure different Accepted for publication Feb 16, 1999. Surgery 1999;126:1-4. Reprint requests: Vic Velanovich, MD, Division of General Surgery, K-8, Henry Ford Hospital, 2799 W Grand Blvd, Detroit, MI 48202-2689. Copyright © 1999 by Mosby, Inc. 0039-6060/99/$8.00 + 0 11/60/97994
domains, depending on their purpose. Domains can be measured in two “dimensions”: objective assessment of functioning or health status and subjective assessment by the patient of his or her own health.2 Although objective measures are attractive to the clinician and researcher because they are more easily quantifiable and reproducible, it is ultimately the improvement of the patient’s perception of their own health and function that is the goal of much of our medical endeavor. The purpose of this article is to review the concepts of quality of life, how it is measured, and the instruments designed to measure it. The focus will be directed at the purposes of quality-of-life instruments, criteria to assess these instruments, types of instruments, and how to choose among these instruments for a given purpose. Throughout I will use the example of gastroesophageal reflux disease (GERD) to illustrate these points. PURPOSES OF QUALITY-OF-LIFE INSTRUMENTS Quality-of-life instruments have a number of applications. Fitzpatrick et al3 have identified 6 categories of applications for quality-of-life instruments. 1. Screening and monitoring for psychosocial problems in individual patient care—directed primarily to patients with chronic disease to measure progress or to identify psychosocial stressors that may require intervention. 2. Population surveys of perceived health problems—use of the instrument as a survey to help health care policy makers in identifying healthrelated quality-of-life concerns as perceived by the general population of a region. SURGERY 1
Surgery July 1999
Table. Examples of quality-of-life instruments Generic
SF-36 Sickness Impact Profile Psychological General Well-Being Scale Nottingham Health Profile McMaster Health Index Questionnaire
Arthritis Impact Measurement Scales Functional Living Index: Cancer Inflammatory Bowel Disease Questionnaire
GERD-HRQL Pain Visual Analogue Scale Gastrointestinal Symptom Rating Scale Dyspepsia Symptom Checklist
Benign Prostatic Hypertrophy Impact Index QL-Index (Cancer)
These are just a few examples of quality-of-life instruments. Clearly there are many more. Information on instruments is readily available through the literature. Additional information can be obtained through the MAPI Research Institute in Lyon, France, or the Medical Outcomes Trust in Boston, Mass.
3. Medical audit—use of quality-of-life measures to provide another outcome endpoint in assessing practice patterns. 4. Outcome measures in health services or evaluation research—the use of quality-of-life measures as endpoints in clinical research or assessment of interventions as traditionally done in academic centers. 5. Clinical trials—the use of quality-of-life endpoints, in addition to traditional endpoints, in clinical trials. 6. Cost-utility analyses—a controversial aspect that involves the use of quality-of-life measures to aid in decisions about the economics of health care.
It is essential for the investigator or clinician to have a clear purpose to choose or develop the right instrument. Given these purposes, quality-of-life instruments can take 3 modes.4 First, quality-of-life instruments can be discriminative, that is, the instrument is used to separate groups of patients depending on quality-of-life results. For example, in GERD it is important to discriminate between patients who are or are not satisfied with the current level of symptoms.5 Instruments can also be predictive, that is, these instruments can be used to predict an outcome, such as return of physical functioning, as a result of the natural history of the disease or a medical intervention. For example, in GERD patients with a low score in the mental health domain of the SF-36 (ie, a higher level of depressive or anxiety symptoms) will have a worse symptomatic outcome after antireflux surgery.6 Last, quality-of-life instruments can be evaluative, that is, the instrument can be used to measure the magnitude of change over time of an individual or group of patients. For example, does medical or surgical intervention improve quality of life in patients with GERD? ESSENTIAL CHARACTERISTICS OF A QUALITY-OF-LIFE INSTRUMENT Quality-of-life instruments have 5 essential characteristics that determine their “quality.”3,7
1. Reliability: the instrument must produce the same result on repeated examinations. 2. Validity: the instrument needs to measure what it was intended to measure. 3. Sensitivity to change (also referred to as responsiveness): the instrument must be able to detect and measure changes in quality of life over time or after an intervention. 4. Appropriateness: the instrument must be appropriate for the health problem and the likely range of effects of the treatment being investigated. 5. Practicality: this characteristic is the ease with which the instrument can be used.
It is important to note that a substantial number of instruments published and in use have not been evaluated for all the above characteristics. Gill and Feinstein8 have noted frequent and significant deficiencies in many quality-of-life instruments published. Therefore it is incumbent on the physician or researcher using these instruments to determine whether they are reliable, valid, responsive, appropriate, and practical instruments for the purpose they intend them to be used. TYPES OF QUALITY-OF-LIFE INSTRUMENTS There are 3 basic types of quality-of-life instruments: generic, disease specific, and symptom severity. The Table provides examples of each. Quality-of-life instruments are essentially questionnaires. The patient answers the questions, referred to as items, about how he or she perceives his or her own health as asked by the question. An aspect or category of quality of life that the instrument measures is called a domain (eg, physical functioning or bodily pain). Items that address the same aspect of quality of life are grouped together so they form the basis of the domain score. Some instruments then summarize these scores into an overall quality-of-life score. Generic instruments. Generic instruments are designed to be broadly applicable across a wide range of types and severity of diseases, across different medical treatments or health interventions, and
Surgery Volume 126, Number 1 across demographic and cultural subgroups.9 Generic instruments measure health-related quality of life in broad terms generally covering the concepts of duration of life, impairments, functional status, perceptions, and social opportunities.9 Not all these categories need be addressed by each instrument, and some instruments are more in depth in their coverage of 1 or more of these domains than others are. Disease-specific instruments. Disease-specific instruments are designed to assess specific diagnostic groups or patient populations, especially with the goal of measuring “clinically important” changes.9 The domains and items chosen for these instruments are directly related to the impairments caused by the disease process or its intervention (for example, sexual functioning in breast cancer patients, activities of daily living in arthritis patients, or dietary habits of patients with GERD). These instruments may produce scores only in each domain measured or they may produce an overall score. Symptom severity instruments. Symptom severity instruments focus only on the symptoms produced by a given disease process without addressing other quality-of-life issues such as social interactions or psychologic stresses. For example, the Gastroesophageal Reflux Disease–Health Related Quality of Life Score measures only heartburn-related symptoms5; it does not measure the impact of GERD on, for instance, social interactions. A symptom as reported by a patient is a verbal expression of subjective experiences that summarize and integrate stimuli from a variety of sources.10 This response includes a pathophysiologic component (which can be measured objectively through clinical, laboratory, or radiologic tests) as well as psychologic and emotional components. Most instruments give an overall score of the symptom severity rather than scores for each component of the symptom. CHOOSING AN INSTRUMENT No one quality-of-life instrument fits all situations. The choice of the instrument cannot be separated from the goals of the investigator or clinician. However, some rules of thumb can be helpful in making the choice.11 In all situations the instrument chosen must be reliable and valid. It must be reliable in the sense that the instrument must produce the same results on repeated examinations under the same conditions, a concept that must be tested for specifically by the developers of the instruments. Therefore ad hoc or “home-grown” instruments that have not been proved reliable (validated) should be viewed with caution. Unfortunately, for example, many of
the questionnaires published that are purported to measure GERD symptoms have not been evaluated for reliability. Validity can be assessed in an informal way, so-called content or “face” validity, or in a formal statistical way, so-called construct validity. Generic instruments, such as the SF-36 survey and the Sickness Impact Profile, have been tested for construct validity. Nevertheless, the content validity may be weak for a given problem.11 Responsiveness (ie, sensitivity to change) is an overriding issue in studies evaluating therapies.12 In this setting the objective is to measure the improvement (or worsening) of quality of life. For example, in GERD the GERD-HRQL symptom severity scale was more responsive to the effects of medical and surgical therapy on symptom severity than the SF-36 survey was.13 However, in studies that seek to assess broad quality-of-life issues associated with the natural history of a disease process responsiveness is not important. Practicality is important. For example, how is the survey administered? By the researcher? By an assistant? Self-administered? In person? Over the telephone or by mail? In the setting of a well-funded clinical trial long and complicated questionnaires that provide maximal data may be appropriate; in contrast, in the setting of a busy clinical practice a short, self-administered questionnaire addressing key issues may be most appropriate. In general, the longer the questionnaire, the more difficult it becomes for a patient to answer completely without assistance. Hence in this situation frequent unanswered items are encountered, lowering the quality of the data. Another question addresses how the scoring will be done. Some generic questionnaire can be computer scored and others are simple enough to be scored easily by hand. Therefore practical issues need to be addressed before an instrument is chosen. Last, and most important, the researcher or clinician needs to have a clear idea of what he or she wishes to determine with the information. If the goal is to determine whether a particular therapy improves symptoms, then a short symptom severity instrument is appropriate. If broader effects within the context of a single disease process are what is required, yet keeping the questionnaire as short as possible, then a disease-specific instrument may be most appropriate. If several disease processes or diseases with wide ramifications are to be studied, then a generic questionnaire may be more appropriate. In some studies a combination of a generic and disease-specific or symptom severity instrument would provide the most information. In summary, quality-of-life data provide pertinent information on patient outcomes. The
Surgery July 1999
researcher will continue to incorporate this endpoint into clinical research and the clinician will need to become competent in the interpretation of this data. REFERENCES 1. Slevin M, Plant H, Lynch D, Drinkwater J, Gregory WM. Who should measure quality of life, the doctor or the patient? Br J Cancer 1988;57:109-12. 2. Testa MA, Simonson DC. Assessment of quality of life outcomes. N Engl J Med 1996;334:835-40. 3. Fitzpatrick R, Fletcher A, Gore S, Jones D, Spiegelhalter D, Cox D. Quality of life measures in health care, I: applications and issues in assessment. BMJ 1992:305:1074-7. 4. Kirshner B, Guyatt G. A methodological framework for assessing health indices. J Chron Dis 1985;38:27-36. 5. Velanovich V, Vallance SR, Gusz JR, Tapia FV, Harkabus MA. Quality of life scale for gastroesophageal reflux disease. J Am Coll Surg 1996;183:217-24. 6. Velanovich V, Karmy-Jones R. Psychiatric disorders affects outcomes of antireflux operations for gastroesophageal
7. 8. 9.
reflux disease. Proceedings of the annual meeting of the Society of American Gastrointestinal Endoscopic Surgeons; 1999 Mar 24-27; San Antonio, Texas. San Antonio: The Society; 1999. Guyatt GH, Feeny DH, Patrick DL. Measuring health-related quality of life. Ann Intern Med 1993;118:622-9. Gill TM, Feinstein AR. A critical appraisal of quality of life measurements. JAMA 1994;272:619-26. Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care 1989;27:S217-32. Wilson IB, Cleary PD. Linking clinical variables with healthrelated quality of life: a conceptual model of patient outcomes. JAMA 1995;273:59-65. Fletcher A, Gore S, Jones D, Fitzpatrick R, Spiegelhalter D, Cox D. Quality of life measures in health care, II: design, analysis and interpretation. BMJ 1992;305:1145-8. Fitzpatrick R, Ziebland S, Jenkinson C, Mowat A, Mowat A. Importance of sensitivity to change as a criterion for selecting health status measures. Qual Health Care 1992;1:89-93. Velanovich V. Comparison of generic (SF-36) vs disease-specific (GERD-HRQL) quality of life scales for gastroesophageal reflux disease. J Gastrointest Surg 1998;2:141-5.
BOUND VOLUMES AVAILABLE TO SUBSCRIBERS Bound volumes of Surgery are available to subscribers (only) for the 1999 issues from the publisher at a cost of $106.00 for domestic, $135.89 for Canadian, and $127.00 for international subscribers for Vol. 125 (January to June) and Vol. 126 (July to December). Shipping charges are included. Each bound volume contains a subject index and an author index, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular journal subscription. Payment must accompany all orders. Contact Mosby, Inc., Subscription Services, 11830 Westline Industrial Dr, St Louis, MO 63146-3318; phone (800)325-4177, ext. 4351, or (314)453-4351.