What are Wound Care Outcomes?

Grant support: Dr. Chen is supported in part by a Mentored Patient Oriented Career Development Award (#K23AR02185-01A1) from the National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Department of Health and Human Services, Bethesda, Maryland, as well as an American Skin Association David Martin Carter Research Scholar Award Introduction How does one measure the impact of pain, dysfunction, and impaired social and psychological well being of a nonhealing leg ulcer? In an era of increased healthcare accountability, how can one document or compare the therapeutic efficacy of treatment in terms of these parameters? In a heterogeneous population, how does one account for the fact that different patients may value the same outcome differently? At the population level, how can policymakers allocate health resources to yield the most benefit per dollar spent for the population? These represent pressing questions for both investigators and consumers of patient-centered research and are of particular relevance to wound care as a specialty. Outcomes research represents a group of specialized methodologies designed to systematically address such questions. Rather than focus on physiologic endpoints (e.g., type of granulation tissue at ulcer base), outcome studies assess the effect of interventions on endpoints that are important to patients: health-related quality of life, functional status, patient satisfaction, cost, quality of care, practice standards/patterns, and patient perspectives on new technology.[1,2] Many wounds represent chronic disease states associated with significant morbidity and resource costs. It is thus crucial for consumers of the wound care literature to be literate in methods of outcome measurement, in particular, measures of clinical efficacy, health-related quality of life, and health economics. Validity, reliability, and responsiveness are concepts readers may use to critically appraise the value of published outcome instruments. This article presents an overview of outcomes in wound care, with the goal of highlighting the importance of outcomes research and of providing readers with skills to better utilize outcomes data to inform their clinical practice. Wound Care Outcomes Outcome measures in wound care may be divided broadly into three categories: 1) measures of clinical efficacy, 2) health-related quality of life, and 3) health economics. Health-related quality of life (HRQoL) is generally assessed using either health status instruments—which may be “generic” or “condition specific”—or by utility assessment (Figure 1). Clinical efficacy measures refer to physiologic or functional endpoints considered significant from a biomedical perspective but that do not directly elicit patients’ view of this endpoint. Proposed efficacy measures in wound care include the percentage of patients healed; mean time to complete healing; percentage change in wound dimensions; percentage of wound area debrided; exudate type and amount; infection reduction rate; and predominant tissue type at wound bed.[2–4] In the context of well-designed, randomized trials, measures of clinical efficacy may inform readers of the true effect of the intervention under study. As a result of their perceived objectivity, clinical efficacy measures have been used widely in therapeutic studies and to justify treatment decisions.[5] Clinical endpoints, however, may represent surrogate outcomes in that they do not necessarily correlate with outcomes that are important to patients, i.e., the way patients feel, function, or the health status changes they value.[6,7] A randomized study of plasmapheresis in rheumatoid arthritis, for example, demonstrated significant reduction in clinical indicators of disease activity, such as the erythrocyte sedimentation rate, but showed no correlation with improvement in patient-important outcomes, such as pain, grip strength, or walking ability.[8] As guides to clinical decision making, exclusive use of clinical efficacy measures may be misleading. HRQoL outcomes address this deficiency by accounting for both the objective presence and the subjective experience of disease.[9] HRQoL can be conceptualized as a construct consisting of four domains: physical function (i.e., one’s sense of physical energy and ability to carry out activities of daily living, such as working and household chores); psychological well being (i.e., one’s psychological response to one’s health, including depression, anxiety, and fear); social function (i.e., one’s ability to engage in meaningful, interpersonal relationships); and somatic sensation (i.e., one’s disease-related symptoms, such as ulcer pain). Two measures to HRQoL are the health status (or psychometric) instrument and the utility assessment. Health status instruments represent questionnaires consisting of multiple items that ask patients to rank their experience of illness in terms of frequency or severity using a quantitative measurement scale (for instance, a 0–5 scale). Items are then scored and combined. Advantages of health status instruments are that they may be used to assess the trade-off for treatments yielding benefits and side effects in different HRQoL domains. Health status instruments may be further subdivided into “generic” or “condition-specific” instruments (Figure 1). Generic instruments capture broad-based HRQoL changes and are therefore applicable to a diversity of disease states. Because of their comprehensive design, generic instruments may be used to directly compare the effect of treatments on HRQoL across disparate disease states. This feature may be useful for policymakers attempting to make funding allocation decisions between competing interventions.[6,10,11] Examples of generic health status instruments that may be useful in wound care studies include the Medical Outcome Study Short-Form 36 (SF-36),[12] Nottingham Health Profile (NHP),[13] and the Sickness Impact Profile (SIP).[11] The main disadvantage of generic instruments is their exclusion of quality-of-life aspects that may be significant for a specific condition, for example, wound-related odor or discomfort related to dressing changes. Generic instruments may thus be unresponsive to small, but clinically significant, changes in HRQoL that may be important determinants of outcome in these illnesses.[10,14] Condition-specific health status instruments, by contrast, focus only on HRQoL aspects pertinent to a particular condition. As a result, they have the advantage of being shorter, more appropriate, and more sensitive to subtle, yet important, changes in HRQoL related to the disease in question.[9,11] These features make them amenable for use in clinical trials.[9] The exquisite responsiveness of condition-specific measures, however, is offset by their lack of generalizability and the consequent inability to make comparison across disease groups.[10] Examples of condition-specific instruments applicable to wound care patients include Skindex,[15] the Hyland New Ulcer Specific Tool,[16] and the Charing Cross Venous Ulcer Questionnaire.[17] Results of both generic and specific health status instruments are summarized as either a “health profile” (a report of the different domain-specific scores) or as a “health index” (a summary score of all domain scores).[18] The major advantage of the health status instrument lies in the level of detail provided about HRQoL. By capturing such detail, health status instruments accurately describe the health state being investigated. Health status instruments, however, do not incorporate the value or preference patients place on the health state or the treatment outcome achieved. The second approach to HRQoL assessment is via utilities. Utilities are quantitative measures of HRQoL that reflect the strength of an individual’s preference or value for a particular health outcome. The utility approach to HRQoL is based on economic and decision theory.[19] Utility scores lie on a continuum from 0 to 1, where 0 represents a preference for a health state equivalent to death and 1 represents a preference for a health state equivalent to perfect health. As an example, consider two hypothetical patients, each with a diabetic foot ulcer measuring 1 inch in diameter. The ulcers are identical. However, the first patient is an image-conscious 16-year-old woman with a physically active peer group, whereas the second patient is an image-indifferent 69-year-old man who lives a sedentary lifestyle. These patients will have different preferences for an identical health state: having a diabetic foot ulcer. The self-conscious 16-year-old woman may give a utility of 0.75 (of 1.00), indicating that the ulcer significantly impairs her HRQoL, whereas the indifferent 69-year-old man may report a utility of 0.98 (of 1.00), indicating he considers his condition comparable to perfect health.[20] The advantage of the utility approach is that it is not simply a description of HRQoL as provided by health status instruments but directly integrates patient preferences into the HRQoL estimate. Utility scores are moreover useful because they can be used to calculate a quantity termed quality-adjusted life year (QALY). This measure recognizes that a year in poor health is not equivalent to a year in full health. It represents a patient’s preference for one year of life at a decreased state of health compared with life less than one year duration at an optimal state of health.[21,22] Investigators generate the QALY by multiplying the patient’s utility score (a percentage score ranging from 0 to 1.0) by their remaining years of life, thereby adjusting the remaining years of a patient’s life by its quality. By combining quality and quantity of life into a single summary score, QALYs may be derived for any disease state. Similar to generic instruments, this feature allows QALYs to be used as a common metric to compare HRQoL effects of interventions across different diseases with the added advantage of directly accounting for patient preferences.[22] Controversy, however, still exists regarding the definition, interpretation, and methods used to derive utility scores, and their conceptual complexity often makes elicitation from patient populations difficult and labor intensive.[10,19] Health Economics Health economic studies are central to outcomes research since all interventions produce health benefit at a cost. Indeed, costs associated with wound care are not insignificant. A study by the Agency for Healthcare Research and Quality (formerly Agency for Health Care Policy and Research) estimated that the national cost of pressure ulcer management exceeded $1.4 billion annually.[23] The average cost to heal one leg ulcer is estimated at $1,951, whereas that for a diabetic foot ulcer is estimated at $29,373.[24,25] Given finite resources, efficient resource allocation is necessary to maximize the health benefit per dollar spent for the wound patient population. Three types of pharmacoeconomic studies consider the trade off between cost and outcome—cost-effectiveness, cost-utility, and cost-benefit analyses. These studies report the efficiency of an intervention in terms of the amount of resources consumed for the health outcome produced.[26] Health economic studies divide costs into “direct” and “indirect” cost. In wound care, direct costs refer to those directly related to wound treatment and may include the cost of primary and secondary dressings and the cost of auxiliary materials (such as saline or tape). Indirect costs refer to costs not directly related to wound care and may include those incurred by wound dressing disposal services, patient days lost from work, and travel costs.[4] The International Committee on Wound Management has published guidelines regarding the types of direct and indirect costs that may be considered in economic analyses in wound care, and the Panel of Cost-Effectiveness in Health and Medicine, convened by the US Public Health Service, has standardized methods to estimate these costs.[27,28] Since the effect of the intervention is as important as the cost, economic analyses also consider the outcome produced for the resources consumed. The choice of outcome measure in the analysis determines whether it is classified as a cost-effectiveness, a cost-utility, or a cost-benefit evaluation. Each type of analysis differs in the type of information it provides. Cost-effectiveness analyses use disease-specific clinical measures as the endpoint, e.g., ulcer healed or wound area debrided. Because the outcome measure is disease specific, cost-effectiveness analysis is useful to compare alternative therapies within a specific disorder, e.g., different dressing types within the disease category of ulcers. For example, dressing A is more cost-effective than dressing B if their cost per unit clinical effects are “$100 per ulcer healed” and “$250 per ulcer healed,” respectively. Cost-effectiveness analysis is not useful, however, for making comparisons between different diseases, because the unit of clinical effect in these cases would not be comparable: one cannot meaningfully compare, for example, the cost-effectiveness of “$100 per ulcer healed” to “$200 per asthma exacerbation cured.”[29] Cost-utility studies circumvent this limitation by using the QALY as the outcome measure. As mentioned, QALYs summarize morbidity and mortality into a summary score that is applicable to the majority of disease states; QALYs may thus be used as a common metric to compare the “cost per QALY gained” across different conditions. A meaningful comparison, for example, may be made on the relative cost utility of ulcer and asthma treatments by examining hypothetical figures for ulcer therapy, “$75 per QALY gained,” and asthma exacerbation therapy, “$100 per QALY gained.” In some cases, however, policymakers may be interested in comparing the relative societal benefits achieved by investing resources in healthcare compared to other sectors, such as education or industry. QALYs are inappropriate denominators in this scenario, as they specifically measure HRQoL. Cost-benefit analyses use the dollar as the metric to compare healthcare with other economic sectors.[30] In cost-benefit analysis, a surplus in benefits relative to costs implies an increase in social welfare, whereas a negative result suggests that social costs would exceed social benefits. The requirement to put a monetary value on health, however, is ethically problematic for some health providers and represents an obstacle to widespread use of cost-benefit analysis.[30] Economic analyses, such as cost effectiveness, cost utility, and cost benefit, represent an additional application for outcome measures that may be important in determining rational resource allocation. Critical Appraisal of Outcome Measures When reading reports purporting to address outcomes, wound care professionals must consider whether the results of the instruments are truly credible. A simple method to assure preliminary credibility is to determine whether these instruments have been tested for three important characteristics: validity, reliability, and responsiveness.[6] Validity refers to the extent to which an instrument measures what it intends to measure. Three methods to ascertain validity are criterion, content, and construct validity. Criterion validity refers to the concept that one may determine the validity of an instrument by comparing its performance against that of an external gold standard. No gold standard exists for the measurement of quality of life, however, and so the validity of these outcome measures must be confirmed using either content or construct validity. Content validity refers to whether the items composing an instrument are deemed sufficiently comprehensive so that the instrument measures what it intends to measure. A panel of experts usually makes this judgment. The degree of agreement between judges is expressed as a content validity index, with a perfect score of 1. The closer the content validity index approaches 1, the more valid the instrument. An acceptable value is > 0.8. Some investigators consider expert opinion insufficient to fully capture patient experience and consequently use patient focus sessions to gather the content required for their instruments. In focus sessions, several patients openly discuss the various ways in which their disease impacts their lives. Focus sessions thus better reflect content validity from the patient perspective. The final type of validity is construct validity, which verifies the performance of the instrument against a theoretical prediction. For example, one would hypothesize that a valid instrument would detect differences between contrasting patient groups, i.e., healthy and sick patients. The instrument is applied to these groups, and the extent to which it may measure group differences is assessed.[6,19] Once validity is shown, the instrument must then demonstrate reliability and responsiveness. Reliability refers to the reproducibility of a measure, i.e., the extent to which it consistently produces the same results under the same circumstances. Reliability is best tested by repeated administrations of the instrument to patient groups whose health status has not changed to confirm that the test results do not change significantly.[10] The most widely used indicators of reliability are the Cronbach’s alpha coefficient of internal consistency and the Pearson’s r correlation coefficient for test-retest and inter-rater reliability.[6] Responsiveness refers to the extent to which an instrument is sensitive to clinically important changes in patients’ health status over time. If an instrument is unresponsive, the study to which it is applied may report a false-negative result: a beneficial treatment may show no apparent difference between treated and untreated patients not because no clinically important difference exists, but because the instrument used was unable to detect such a difference.[5,6] If trials do not cite studies demonstrating that their outcome measures have been tested for these characteristics, readers must be cautious when interpreting the outcomes data. Conclusion A diversity of outcome measures is available to assess the effectiveness of treatments in wound care. Measures of clinical efficacy are relatively objective measures of physiologic endpoints that may not necessarily reflect outcomes that are important to patients. HRQoL, by contrast, attempts to capture patient-important outcomes, and may be measured using either health status instruments or utility assessment. Clinical efficacy and utilities may also be incorporated into health economic analyses, such as cost-effectiveness, cost-utility, and cost-benefit analysis, which may be important in rational resource allocation. The choice of which outcome measure or economic analysis to use depends on the particular question one is attempting to answer. In general, when trying to compare the effect of interventions on an outcome within a disease, all measures are applicable, although condition-specific health instruments offer the advantage of highlighting disease-specific features that may be important determinants of HRQoL for the condition in question. When trying to compare outcomes across disparate conditions, however, only generic instruments or utility scores are appropriate. Utility assessment further incorporates patient preferences into the HRQoL estimate. In all cases, readers must ensure adequate testing for validity, reliability, and responsiveness; otherwise, the data may be potentially biased and thus difficult to interpret. A better understanding of the strengths and drawbacks of various outcome and economic analytic methods may improve the quality and efficiency of wound care at the patient and at the population level.