Rationale: Although exacerbations are an important problem in chronic obstructive pulmonary disease (COPD) and a target of intervention, there is no valid, standardized tool for assessing their frequency, severity, and duration.
Objectives: This study tested the properties of the Exacerbations of Chronic Pulmonary Disease Tool (EXACT), a new patient-reported outcome diary.
Methods: A prospective, two-group, observational study was conducted in patients with COPD. The acute group (n = 222) was enrolled during a clinic visit for exacerbation with follow-up visits on Days 10, 29, and 60. The stable group (n = 188), recruited by telephone or during routine visits, was exacerbation free for at least 60 days.
Measurements and Main Results: Acute patients completed the diary on Days 1–29 and 60–67; stable patients for 7 days. All patients provided stable-state spirometry and completed the St. George Respiratory Questionnaire-COPD (SGRQ-C). Acute patient assessments included clinician and patient global ratings of exacerbation severity and recovery. Mean age of the sample (n = 410) was 65 (± 10) years; 48% were male; stable FEV1 was 51% predicted (± 20). Internal consistency (Pearson separation index) for the EXACT was 0.92, 1-week reproducibility (stable patients; intraclass correlation) was 0.77. EXACT scores correlated with SGRQ-C (r = 0.64; P < 0.0001) and differentiated acute and stable patients (P < 0.0001). In acute patients, scores improved over time (P < 0.0001) and differentiated between degrees of clinician-rated exacerbation severity (P < 0.05). EXACT change scores differentiated responders and nonresponders on Day 10, as judged by clinicians or patients (P < 0.0001).
Conclusions: Results suggest the EXACT is reliable, valid, and sensitive to change with exacerbation recovery.
The Exacerbations of Chronic Pulmonary Disease Tool is a patient-reported outcome measure designed to provide researchers with a single, standardized, reliable, and valid method for assessing frequency, severity, and duration of exacerbations of chronic obstructive pulmonary disease.
Prior work established the content validity of the instrument. Results of this first quantitative validation study indicate this 14-item daily diary is reliable, valid, and sensitive to changes during recovery from exacerbation.
Historically, counts of unscheduled clinic or emergency room visits or hospitalizations have been used as a metric for exacerbation frequency, with clinic and emergency room visits rated “moderate” and those requiring hospitalizations rated “severe” (6, 10, 11). However, event-based definitions are indirect measures of exacerbation, and their validity is unclear because of international differences in health care coverage and the fact that as many as 50–70% of exacerbations are unreported (4, 12, 13).
Patient diary cards are direct measures of the signs or symptoms of exacerbation. Advantages to this approach include standardized data, reduced recall bias, and the potential to identify events and determine resolution based on a predefined scoring algorithm. To date, diary cards have varied in content and scoring. Most cards include such symptoms as dyspnea and cough, and such signs as sputum volume. Some also include sore throat, nasal discharge, sputum purulence, and fever (4, 6, 14–18). Response options also vary and include instructing patients to rate the severity of symptoms directly or asking patients if their symptoms had changed that day beyond their usual variability (yes or no). The empirical support for scoring diary cards is limited. For example, the Anthonisen definition and rating system based on the number of “major” and “minor” respiratory symptoms is commonly used despite the absence of data on its validity (19).
Most diary cards are used to quantify frequency of exacerbation. Less attention has been given to severity and patterns of recovery, with the latter particularly important for acute treatment trials. From a clinical trial perspective, none of the instruments meet current standards for patient-reported outcome (PRO) measures in pharmaceutical development trials (20), including the application of systematic methods for optimizing content validity and explicit testing for reliability and validity. Use of a single, standardized, validated measure of exacerbations of COPD will improve the understanding of the disease and its treatment by reducing random error and misclassification bias, and facilitating comparison and aggregation of results across studies to better assess benefits and risks of interventions.
This paper reports the first validation of a new instrument, the EXAcerbations of Chronic Pulmonary Disease Tool (EXACT). Details of its development are presented elsewhere (21). This paper presents the methods and results of the first quantitative test of the instrument's reliability and validity, including its sensitivity to improvement in patients with a medically confirmed exacerbation of COPD. The results of this study have been previously reported in the form of abstracts (22, 23).
This prospective observational study recruited patients with COPD from 38 medical clinics in 20 states of the United States. Inclusion criteria were as follows: medical diagnosis of COPD or chronic bronchitis, post-bronchodilator FEV1 less than 80% and FEV1/FVC less than 70%, at least 40 years of age, and a smoking history of at least 10 pack-years. Patients with comorbid asthma or clinically relevant bronchiectasis were excluded.
To assess the reproducibility and variability of EXACT scores in stable conditions, test the ability of scores to differentiate stable from acutely ill patients, evaluate sensitivity to change during recovery from exacerbation, and evaluate convergent and known-groups validity, two groups of patients were enrolled: acute and stable. Patients in the acute group were enrolled during a clinic visit for an exacerbation of COPD. Visits on Days 1, 10 (± 2), 29 (± 2), and 60 (± 7) included clinician assessments and patient-completed questionnaires. Patients also completed a diary Days 1–29 (± 2) (acute state) and 60–67 (clinically stable state) using a personal digital assistant (Palm Tungsten E; Palm, Inc., Sunnyvale, CA) (see the online supplement for additional e-diary information).
Patients in the stable group, recruited during routine visits or by telephone, had a history of at least one exacerbation in the previous 24 months with no clinic, emergency room, or hospital visit for exacerbation during the previous 60 days. Clinician assessment and patient questionnaires were completed on Day 1 and patients completed the e-diary on Days 1–7.
The protocol was approved by the appropriate institutional review boards and written consent was obtained from participants before the initiation of data collection.
Questionnaires completed during clinic visits included the St. George Respiratory Questionnaire-COPD (SGRQ-C), a 40-item version of the SGRQ (24, 25). Patients in the acute group also provided global ratings of exacerbation severity (mild, moderate, severe, or very severe) at enrollment and change at subsequent visits (much worse, worse, no change, better, or much better).
The 27-item e-diary contained the EXACT item pool (23 items); rescue medication use; clinic contact; and global ratings of health status and change. This analysis concerns the final 14 items that form the EXACT, after an extensive process of item reduction (26) (see online supplement).
The EXACT diary assesses breathlessness, cough and sputum, chest symptoms, difficulty bringing up sputum, feeling tired or weak, sleep disturbance, and feeling scared or worried about their condition. Each attribute or item is assessed on a five- or six-point ordinal scale and summed to yield a total score that is converted to a 0–100 scale, with higher scores indicating a more severe health state or exacerbation. Three respiratory symptom domains are embedded in the instrument (breathlessness, cough and sputum, and chest symptoms), each scored on a 0–100 scale with higher scores indicating more severe symptoms.
Clinicians rated patient dyspnea on enrollment using the Modified Medical Research Council (MMRC) scale (27). For acute patients, physicians provided global ratings of exacerbation severity at enrollment (mild, moderate, severe, or very severe) and change at subsequent visits (much worse, worse, no change, better, much better, or resolved). The most recent stable state spirometry results from the past 12 months were obtained from the medical record or, if not on record, were obtained when the patient was clinically stable.
Details of the statistical analyses are provided in the online supplement. Demographic and clinical characteristics of the acute and stable groups were tested for comparability. Reliability was evaluated in terms of internal consistency and reproducibility. Validity was assessed by examining the relationship among EXACT scores and SGRQ-C, MMRC, rescue medication use, stable-state FEV1% predicted, and GOLD stage. The ability of scores to differentiate stable and acute patients (known-groups validity) and to detect change over time in the acute patients over the first 7 days of the event (sensitivity to change) were both tested simultaneously, controlling for age, comorbidity status (<1 or ≥2), and stable-state FEV1. The ability of scores to differentiate between patients with different levels of clinician-rated exacerbation severity and detect change over 27 days of the event were both tested simultaneously using a similar statistical model. Finally, responsiveness was tested by comparing EXACT scores of patients categorized as “responders” and “nonresponders” on Day 10, with responder group assignment based on clinician and patient global assessments during the Day 10 visit. Individual plots were examined to assess patterns of change over time indicative of persistence, recovery, and relapse.
To assist with interpretation, triangulation analyses were performed (28), stratifying mean EXACT change scores from Days 1–27 by global ratings of change and juxtaposed with percent change, statistical effect size (ES), and SGRQ-C change score. To place these scores in perspective, day-to-day intraindividual variability in EXACT scores during a stable state (Days 1–7 in the stable group) and during recovery from exacerbation (Days 1–27 and 60–67 in the acute group) were determined, expressed as the mean within-subject standard deviation over time, and summarized using the coefficient of variation (CV). Differences in intraindividual variability by GOLD status in stable patients and over time during recovery in acute patients were tested using analysis of variance.
All available data were used in the analyses with no imputation.
Demographic and clinical characteristics are shown in Table 1. The stable group reported a longer history of COPD and fewer exacerbations in the previous year. As expected, SGRQ-C scores on enrollment were significantly higher in the acute group, reflecting their poorer health status. Validation analyses for Day 1 data were based on the entire sample (n = 410). The personal digital assistant compliance rate (actual divided by expected entries) was 94 ± 12% for the stable group Days 1–7 and 92 ± 14 for the acute group Days 1–27. Most of the patients in the acute group (196; 88%) returned for the additional Days 60–67 assessment.
Characteristic | Acute Group (n = 222) | Stable Group (n = 188) | P Values | |||
---|---|---|---|---|---|---|
Demographic | ||||||
Age, mean (SD) | 65.1 (10.5) | 65.8 (9.7) | 0.48 | |||
Sex, male, n (%) | 104 (47) | 91 (48) | 0.75 | |||
Race/ethnicity, n (%)* | ||||||
White | 207 (93) | 166 (88) | 0.08 | |||
Employment status, n (%) | ||||||
Retired | 117 (53) | 110 (59) | 0.24 | |||
Clinical | ||||||
COPD diagnosis, yr, mean (SD) | 6.4 (4.9) | 7.8 (6.7) | 0.02 | |||
Chronic bronchitis, n (%) | 65 (29) | 51 (27) | 0.63 | |||
Current smoker, n (%) | 71 (32) | 58 (31) | 0.81 | |||
Exacerbation HX, prior 12 mo | ||||||
Number, mean (SD) | 2.2 (2.1) | 1.3 (1.2) | <0.0001 | |||
Stable spirometry, mean (SD)† | ||||||
FEV1, L | 1.4 (0.6) | 1.3 (0.6) | 0.88 | |||
FEV1% predicted | 50.5 (20.4) | 51.2 (19.6) | 0.75 | |||
FVC, L | 2.5 (0.9) | 2.4 (0.9) | 0.72 | |||
FEF25–75%, L | 1.1 (1.6) | 1.1 (1.3) | 0.73 | |||
SGRQ-C, mean (SD)‡ | ||||||
Total | 62.1 (17.7) | 50.8 (19.9) | <0.0001 | |||
Symptom | 70.1 (18.7) | 60.8 (21.2) | <0.0001 | |||
Activity | 76.3 (21.7) | 69.7 (23.5) | 0.0035 | |||
Impacts | 51 (20.5) | 36.8 (22.3) | <0.0001 | |||
EXACT scores Day 1, mean (SD)§ | ||||||
Total score | 47.9 (9.6) | 36.1 (12.7) | <0.0001 | |||
Breathlessness | 48.9 (16.1) | 35.6 (17.9) | <0.0001 | |||
Cough and sputum | 48.3 (19.8) | 33.3 (21.9) | <0.0001 | |||
Chest symptoms | 42.6 (19.6) | 23.9 (18.3) | <0.0001 | |||
EXACT scores Day 60, mean (SD)§ | ||||||
Total score | 37.3 (14) | — | — | |||
Breathlessness | 38.6 (20.4) | — | — | |||
Cough and sputum | 32.2 (20.7) | — | — | |||
Chest symptoms | 26.3 (21.5) | — | — |
Reliability estimates for EXACT scores are shown in Table 2. Internal consistency levels were high and scores were reproducible over the 1-week test–retest period with a mean difference of less than one point on the 100-point scale.
Internal Consistency (n = 410) | Reproducibility (Days 1–7) (n = 171)* | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
EXACT† | Cronbach α | ICC | Day 1 Mean (± SD) | Day 7 Mean (± SD) | Mean Difference | Effect Size | ||||
Total score | 0.91‡ | 0.77 | 36 (13) | 36 (14) | −0.35 | 0.03 | ||||
Breathlessness | 0.87 | 0.71 | 36 (18) | 36 (21) | 0.43 | 0.04 | ||||
Cough and sputum | 0.69 | 0.65 | 33 (22) | 32 (19) | −0.68 | 0.03 | ||||
Chest symptoms | 0.87 | 0.64 | 24 (18) | 25 (20) | 0.74 | 0.02 |
EXACT scores were significantly related to health status on enrollment, as measured by the SGRQ-C (Table 3). The strongest relationship between the EXACT and clinical indicators of airway obstruction (FEV1% predicted, MMRC, and rescue medication use) was with the breathlessness domain. The multivariate statistical model supported the known-groups validity of the instrument, with EXACT scores differentiating stable from acute patients, including those rated by clinicians as mild, moderate, and severe (F = 46.0; P < 0.001). Mean EXACT total (± SD) scores for these four groups at enrollment were as follows: stable = 36 ± 13 (n = 188); mild = 43 ± 9 (n = 52); moderate = 48 ± 9 (n = 133); and severe = 53 ± 9 (n = 37). All pair-wise comparisons were significant (P < 0.05) with the exception of moderate versus severe.
FEV1% Predicted (n = 407)† | MMRC (n = 410) | Rescue Medication Use (n = 410)‡ | SGRQ-C | |||||||
---|---|---|---|---|---|---|---|---|---|---|
EXACT | Total (n = 395) | Symptoms (n = 409) | Activity (n = 402) | Impacts (n = 396) | ||||||
Total score | −0.07 | 0.30§ | 0.37§ | 0.64§ | 0.56§ | 0.46§ | 0.62§ | |||
Breathlessness | −0.19‖ | 0.40§ | 0.36§ | 0.66§ | 0.47§ | 0.62§ | 0.58§ | |||
Cough and sputum | 0.00 | 0.06 | 0.26§ | 0.35§ | 0.44§ | 0.14¶ | 0.37§ | |||
Chest symptoms | 0.00 | 0.16¶ | 0.27§ | 0.46§ | 0.49§ | 0.24§ | 0.47§ |
Known-groups validity and sensitivity to change were also supported in the repeated-measure analysis of covariance for Days 1–7 in acute and stable patients (Figure 1A). In acute patients, sensitivity to change was shown over Days 1–27 (Figure 1B) with known-groups validity supported using clinician rating of exacerbation severity. In the latter case, similar results were found when patient severity rating was used in the model (χ2 = 4392; P < 0.0001), with effects for time (F = 321; P < 0.0001); severity (F = 3.47; P < 0.05); and age (F = 4.23; P < 0.05).

Figure 1. Mean (± SE) daily Exacerbations of Chronic Pulmonary Disease Tool (EXACT) total scores over time. (A) Mean (± SE) EXACT scores Days 1–7 in acute and stable patients. Results of repeated-measures analysis of covariance were as follows: Model: χ2 = 2298; P < 0.0001. Significant effects were found for group (acute versus stable; F = 49.90; P < 0.0001) and time (F = 94.01; P < 0.0001.) Significant covariates: age (F = 8.84; P < 0.01) and comorbidity (0–1; ≥2) (F = 5.66; P < 0.05); stable FEV1 was not significant (F = 0.91; P = 0.34). (B) Mean (± SE) EXACT scores Days 1–27 of an exacerbation (acute group). Results of repeated-measures analysis of covariance were as follows: Model: χ2 = 4465; P < 0.0001. EXACT score improvement over time was statistically significant (F = 335; P < 0.0001), as were score differences by clinician global rating of exacerbation severity (F = 4.34; P < 0.05) (data not shown). Results were similar when patient severity rating was used (χ2 = 4392; P < 0.0001), with effects for time (F = 321; P < 0.0001); severity (F = 3.47; P < 0.05); and age (F = 4.23; P < 0.05).
[More] [Minimize]For the responder analyses, significant differences in EXACT scores were found between responders and nonresponders on Day 10 (P < 0.0001), whether judged by clinician or patient (Table 4). Mean change for responders exceeded 10 points, whereas nonresponders averaged less than 3.5 points. Similar results were found when the data were stratified by stable-state FEV1, comparing groups above and below the median value (data not shown).
Global Rating of Change - Source | Responders | Nonresponders | F Statistic* | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Day 1 Mean (SD) | Day 10 Mean (SD) | Change Mean (SD) | Day 1 Mean (SD) | Day 10 Mean (SD) | Change Mean (SD) | ||||||
Clinician rating† | 47.6 (9.2) | 37.1 (12.9) | −10.6 (11.7) | 49.1 (10.5) | 45.7 (10) | −3.4 (11) | 23.82‡ | ||||
Patient rating§ | 48.4 (9.2) | 35.8 (13) | −12.6 (11) | 47.5 (10.2) | 45.5 (9.7) | −2 (10.6) | 39.49‡ | ||||
Convergence‖ | 48.5 (9) | 35.7 (13.3) | −12.8 (11) | 49.5 (10.6) | 46.9 (10.2) | −2.6 (11.2) | 22.90‡ |
Triangulation results for group-level data showed that clinician global ratings of “better” at Day 29 of an exacerbation (n = 82) were associated with a group mean EXACT score change of 7.3 ± 9.7, representing a 14% score improvement, a statistical ES of 0.80, and corresponding to a change in SGRQ-C of −5.4 ± 12.6, which exceeds the SGRQ-C minimal important difference (MID) of 4. Mean scores for patients whose exacerbation was rated as “much better” (n = 48) were 10.6 ± 12.6, a 21% change, an ES greater than 1, and an SGRQ-C score of −8.8 ± 12.5; for those rated completely resolved (n = 34), mean score was 17.9 ± 14.5, a 36% improvement with an SGRQ-C change of −9.6 ± 18.1. Similar results were seen with patient global ratings of change.
Mean intraindividual variability of EXACT scores for patients in the stable group from Days 1–7 (n = 188) was 5.2 (2.7), a CV of 52%, with no differences by GOLD stage (F = 0.99; P = 0.40). Mean within-subject variability during recovery from an exacerbation (acute group, Days 1–27) was 6.8 ± 3.1, a CV of 46%. Day-to-day variability was significantly higher on Days 1–5 (5.5 ± 3) than on Days 11–15 (4.9 ± 3.5), Days 16–20 (4.5 ± 3.1), and Days 21–28 (4.7 ± 3.1) (P < 0.05). Intraindividual variability on Days 60–67 for the acute group was 4.7 ± 3.1 with a CV of 66% (n = 196). Further information on individual variability in EXACT scores is provided in the online supplement.
The EXACT-PRO Initiative was formed to address the need for a standard, validated instrument for quantifying the frequency, severity, and duration of exacerbations of COPD in clinical trials. Prior work established the content validity of the EXACT. This study tested the performance characteristics of the instrument in stable patients with a history of exacerbations of COPD and acutely ill patients with a medically confirmed exacerbation. Scores were found to be reliable, with internal consistency estimates for the total score exceeding the recommended value of 0.90 for measures used for decision making in applied settings (29). Scores were also reproducible over 1 week in stable patients, with the mean score difference effectively zero. The relationship between EXACT scores and criterion measures was in the predicted direction, with the strongest relationship observed between health status (SGRG-C) and EXACT total score. As expected, direct and indirect indicators of dyspnea (MMRC and rescue medication use) and airway obstruction (FEV1 % predicted) were more strongly related to the EXACT breathlessness domain score.
This study was designed to address two critical attributes of the instrument before proceeding with further testing in larger prospective studies: its ability to differentiate stable from acutely ill patients with exacerbation and its ability to detect improvements in patients during recovery from a medically confirmed exacerbation. Results showed that EXACT scores differentiated stable patients from those who were exacerbating. In the absence of a reference standard for grading the severity of exacerbations of COPD, we used a clinician global assessment of severity, from mild to very severe. EXACT scores differentiated patients who were stable from those with clinician-judged mild events and patients with mild and moderate exacerbations. Group-level variance within each of the severity strata may be a function of score error, true patient-to-patient variability, within- or between-clinician variability in their appraisal of patients, or a combination of these factors. The mean difference between moderate and severe events was in the predicted direction but did not reach statistical significance, perhaps because of the absence of the most severe patients (i.e., those requiring treatment in the emergency room or hospital). The highest recorded Day 1 score in this sample was 75, indicating sufficient room in the upper end of the scale to record more severe events. Studies using the EXACT to assess exacerbations in patients admitted to emergency rooms or hospital are warranted, with specific attention given to score magnitude during these acute events and patterns of change with recovery.
When patients were categorized in terms of their response to initial therapy for exacerbation on Day 10, EXACT scores clearly discriminated between those patients judged by clinicians to have improved (responders) and those who had not (nonresponders), supporting the sensitivity of the instrument to improvement during recovery from exacerbations. Once again, in the absence of a reference standard for determining recovery from an exacerbation of COPD, we used a clinician global assessment of change on a six-point scale (much worse, worse, no change, better, much better, or resolved), complemented by patient global ratings. The variance in the two responder groups may be a function of instrument error; the combination rule for determining responder group (three levels of change assigned to each group); true patient variability; or within- or between-clinician variability in the appraisal of patients, which is based on patient interview, recollection of the patient's baseline state, and clinical experience.
As a group, the pattern of recovery was monotonic reaching a stable level at approximately 12 days. As one might expect, individual recovery patterns showed patient-to-patient variability, with some discrepancy between EXACT score change and responder grouping. One explanation for this discrepancy is the difference in assessment methodology. The EXACT diary provides standardized contemporaneous patient-reported symptoms, whereas the clinician assessment is based on observation supplemented by patient and clinician recall of the patient's symptoms and health during the previous visit. It is possible that the discrepancy would be smaller if clinical assessments were standardized. A second explanation may lie in the use of global assessments of change to determine the therapeutic course of action. Ratings of “improved” or “resolved” may indicate that the patient's condition is improving, and likely to continue to improve, so that no change in therapy is warranted. In these cases, the EXACT score would continue to improve as the patient recovers from the event. Further study of EXACT score changes with exacerbations and the factors contributing to clinician global assessment decisions would shed important light on both assessment methods.
Several different threshold values are required to use the EXACT, including values indicating onset and recovery of an event in individual patients, and MIDs for frequency (counts), severity (score), and duration (days) for hypothesis testing in groups of patients. This study was not designed to capture the onset of an event or assess score differences in patients who were treated in an emergency department or hospital. With these limitations in mind, interpretation of the instrument at the individual level begins with an analysis of the available data and insight gained at the group level. The average day-to-day within-patient variability in a stable state was five points, supporting the use of a 5–7 day baseline period to establish individual baseline values and suggesting that a 1-day deterioration or improvement of five points is not meaningful in determining onset or recovery from an event. Stable patients reported an average score of 36 points, whereas the acute group mean was 48 points, a 12-point difference. By Day 29, acute patients rated by clinicians as “much better” had improved an average of 11 points, whereas those considered clinically resolved showed an average improvement of 18 points, both changes associated with very large statistical ESs and improvements in health status. Based on these results and further analyses, a preliminary scoring algorithm was developed to identify events in individual patients and determine when an event is over and a new event begins. Prospective studies are underway to test these algorithms to compute event frequency and duration at the individual patient level.
In terms of a MID for frequency, Calverley's (30) assessment of treatment effects across multiple large trials suggested a 20–25% reduction in the frequency of reported exacerbations in patients with an FEV1 less than 50% predicted is meaningful. The EXACT should complement frequency counts based on health care use by showing the effect of treatment on unreported events. The extent to which Calverley's proposed MID translates into frequency of unreported exacerbations, mild versus moderate or severe events, and across a broader range of disease severity warrants further study. In reference to duration, a reasonable starting point is to assume that a treatment that reduces the duration of an exacerbation by 1 day would be meaningful to patients. The extent to which event frequency, duration, and severity translate prospectively into reduced morbidity and mortality is an empirical question worth pursuing.
The EXACT was developed to detect and quantify exacerbations of COPD. It was not designed as a diagnostic tool to differentiate exacerbations from other medical problems, such as decompensated congestive heart failure (CHF) or pulmonary embolism (PE). Acute events associated with admission to the clinic, emergency room, or hospital would involve a medical evaluation with a differential diagnosis to rule out other causes of the event. In a clinical study, utilization events attributed to CHF or PE would be excluded from a frequency count of COPD exacerbations. Results of this study suggest the EXACT will pick up unreported episodes in which patients experience an increase in dyspnea, cough, sputum, and systemic symptoms consistent with the clinical definition of exacerbations of COPD. In the case of randomized controlled trials, patients with heart failure would either be excluded from participation, which is often the case, or the non-COPD events would be distributed randomly across treatment groups unless the treatment has a therapeutic effect on CHF or PE. Researchers would be wise to examine patterns of EXACT scores within and across patients with a history or recent diagnosis of CHF to determine the likelihood that unreported events may have been caused by a brief episode of cardiac decompensation rather than COPD per se.
Although the instrument was developed specifically for use as an outcome measure in clinical trials, it is possible that in the future the measure may be useful for monitoring patients in clinical practice. Initially, this may be limited to standardizing the assessment of patient symptoms during telephone or electronic screening for exacerbations or during clinic visits. With further validation and advances in electronic media, one can imagine a time when patients might use the EXACT as part of an early warning electronic diary system that detects change and prompts the patient to take prescribed self-care actions or notify a clinician for further evaluation.
Results of this study showed that EXACT scores are internally consistent and reproducible, valid, and sensitive to changes that occur during recovery from exacerbation. The instrument is ready for further study in natural history studies and acute treatment and prevention trials.
Members of the EXACT-PRO Study Group contributed to the design of the study presented here and the interpretation of results but did not participate in the analysis of the data or the writing of this paper. The following companies have provided unrestricted funds to support the EXACT-PRO Initiative: AstraZeneca, GlaxoSmithKline, Pfizer Inc., Boehringer Ingelheim, Merck & Company, Sepracor Inc., Forest Laboratories, Inc., Novartis, Schering-Plough, Adams Respiratory Therapeutics, Bayer, Atlanta AG (Nycomed), and Ortho-McNeil. The data analysis for this paper was generated using SAS/BASE and SAS/STAT software, Version 9.1.3 of the SAS System for PC, SAS Institute Inc., Cary, North Carolina. The UBC EXACT-PRO Initiative Team includes: Wen-Hung Chen, Ph.D.; Laurie Roberts, M.P.H.; Randall Winnette, B.S.; Lindsey Murray, B.A.; Ren Yu, M.A.; Kellee Howard, M.A.; Jennifer Petrillo, B.S.; Charlotte Cates, M.A.; and Chris Thompson, B.S. The authors gratefully acknowledge individuals from the US Food and Drug Administration for their useful comments throughout the instrument development process.
1. | Anzueto A, Leimer I, Kesten S. Impact of frequency of COPD exacerbations on pulmonary function, health status and clinical outcomes. Int J Chron Obstruct Pulmon Dis 2009;4:245–251. |
2. | Halbert RJ, Isonaka S, George D, Iqbal A. Interpreting COPD prevalence estimates: what is the true burden of disease? Chest 2003;123:1684–1692. |
3. | Mannino DM, Braman S. The epidemiology and economics of chronic obstructive pulmonary disease. Proc Am Thorac Soc 2007;4:502–506. |
4. | Seemungal TA, Donaldson GC, Paul EA, Bestall JC, Jeffries DJ, Wedzicha JA. Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1418–1422. |
5. | Seemungal TA, Hurst JR, Wedzicha JA. Exacerbation rate, health status and mortality in COPD: a review of potential interventions. Int J Chron Obstruct Pulmon Dis 2009;4:203–223. |
6. | Vijayasaratha K, Stockley RA. Reported and unreported exacerbations of COPD: analysis by diary cards. Chest 2008;133:34–41. |
7. | Bach PB, Brown C, Gelfand SE, McCrory DC. Management of acute exacerbations of chronic obstructive pulmonary disease: a summary and appraisal of published evidence. Ann Intern Med 2001;134:600–620. |
8. | Metlay JP, Singer DE. Outcomes in lower respiratory tract infections and the impact of antimicrobial drug resistance. Clin Microbiol Infect 2002;8:1–11. |
9. | Cazzola M, MacNee W, Martinez FJ, Rabe KF, Franciosi LG, Barnes PJ, Brusasco V, Burge PS, Calverley PM, Celli BR, et al. Outcomes for COPD pharmacological trials: from lung function to biomarkers. Eur Respir J 2008;31:416–469. |
10. | Rodriguez-Roisin R. Toward a consensus definition for COPD exacerbations. Chest 2000;117:398S–401S. |
11. | Snider GL. Nosology for our day: its application to chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2003;167:678–683. |
12. | Langsetmo L, Platt RW, Ernst P, Bourbeau J. Underreporting exacerbation of chronic obstructive pulmonary disease in a longitudinal cohort. Am J Respir Crit Care Med 2008;177:396–401. |
13. | Rennard SI, Leidy NK. Definition and severity of COPD exacerbations. In: Wedzicha W, Martinez F, editors. Exacerbations of chronic obstructive pulmonary disease (COPD). New York: Informa Healthcare; 2009. pp. 1–14. |
14. | Burge S, Wedzicha JA. COPD exacerbations: definitions and classifications. Eur Respir J Suppl 2003;41:46s–53s. |
15. | Calverley PM, Rennard S, Nelson HS, Karpel JP, Abbate EH, Stryszak P, Staudinger H. One-year treatment with mometasone furoate in chronic obstructive pulmonary disease. Respir Res 2008;9:73. |
16. | Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004;23:932–946. |
17. | Lorenz J, Busch W, Thate-Waschke IM. Moxifloxacin in acute exacerbations of chronic bronchitis: clinical evaluation and assessment by patients. J Int Med Res 2001;29:61–73. |
18. | Seemungal TA, Donaldson GC, Bhowmik A, Jeffries DJ, Wedzicha JA. Time course and recovery of exacerbations in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2000;161:1608–1613. |
19. | Anthonisen NR, Manfreda J, Warren CP, Hershfield ES, Harding GK, Nelson NA. Antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Ann Intern Med 1987;106:196–204. |
20. | Food and Drug Administration (FDA). Guidance for industry on patient-reported outcome measures: use in medical product development to support labeling claims. Fed Regist 2009;74:65132–65133. |
21. | Leidy NK, Wilcox T, Jones PW, Murray L, Winnette R, Howard K, Petrillo J, Powers J, Sethi S, for the EXACT-PRO Study Group. Development of the exacerbations of chronic obstructive pulmonary disease tool (EXACT): a patient-reported outcome measure. Value Health 2010;13:965–975. |
22. | Leidy NK, Wilcox T, Howard K, Petrillo J, Jones PW, Sethi S, Powers J, for the EXACT-PRO Study Group. Reliability and validity of the EXAcerbations of Chronic Pulmonary Disease Tool (EXACT): a new outcome measure for evaluating exacerbations in COPD. Poster presented at the American Thoracic Society International Conference. May 2008, Toronto, Canada. Poster board #D45. Am J Respir Crit Care Med 2008;177:A139. |
23. | Leidy NK, Wilcox TK, Sethi S, Jones PW, the EXACT-PRO Study Group. How stable is stable in COPD? An analysis of day-to-day exact score variability in acute and stable patients with COPD. Poster presented at the American Thoracic Society International Conference. May 2009, San Diego, CA. Poster board #K77. Am J Respir Crit Care Med 2009;179:A1525. |
24. | Jones PW, Quirk FH, Baveystock CM, Littlejohns P. A self-complete measure of health status for chronic airflow limitation. The St. George's respiratory questionnaire. Am Rev Respir Dis 1992;145:1321–1327. |
25. | Meguro M, Barley EA, Spencer S, Jones PW. Development and validation of an improved, COPD-specific version of the St. George respiratory questionnaire. Chest 2007;132:456–463. |
26. | Jones PW, Chen W-H, Wilcox TK, Sethi S, Leidy NK, for the EXACT-PRO Study Group. Characterizing and quantifying the symptomatic features of COPD exacerbations. Chest (In press) |
27. | Fletcher CM, Elmes PC, Fairbairn AS, Wood CH. The significance of respiratory symptoms and the diagnosis of chronic bronchitis in a working population. BMJ 1959;2:257–266. |
28. | Leidy NK, Wyrwich KW. Bridging the gap: using triangulation methodology to estimate minimal clinically important differences (MCIDs). COPD 2005;2:157–165. |
29. | Nunnally JC, Bernstein IH. Psychometric theory, 3rd ed. New York: McGraw-Hill; 1994. |
30. | Calverley PM. Minimal clinically important difference: exacerbations of COPD. COPD 2005;2:143–148. |