0
Clinical Investigations: COPD |

Using Quality of Life to Predict Hospitalization and Mortality in Patients With Obstructive Lung Diseases* FREE TO VIEW

Vincent S. Fan, MD, MPH; J. Randall Curtis, MD, MPH, FCCP; Shin-Ping Tu, MD, MPH; Mary B. McDonell, MS; Stephan D. Fihn, MD, MPH; for the Ambulatory Care Quality Improvement Project Investigators
Author and Funding Information

*From the Health Services Research and Development Center of Excellence (Drs. Fan and Fihn, and Ms. McDonell), VA Puget Sound Health Care System, Seattle, WA; and the Department of Medicine (Drs. Curtis and Tu), University of Washington, Seattle, WA.

Correspondence to: Vincent S. Fan, MD, HSR&D (152), VA Puget Sound Health Care System, 1660 S Columbian Way, Seattle, WA 98108-1597; e-mail: vfan@u.washington.edu


*From the Health Services Research and Development Center of Excellence (Drs. Fan and Fihn, and Ms. McDonell), VA Puget Sound Health Care System, Seattle, WA; and the Department of Medicine (Drs. Curtis and Tu), University of Washington, Seattle, WA.


Chest. 2002;122(2):429-436. doi:10.1378/chest.122.2.429
Text Size: A A A
Published online

Study objectives: Condition-specific measures of quality of life (QOL) for patients with COPD have been demonstrated to be highly reliable and valid, but they have not conclusively been shown to predict hospitalization or death.

Objective: We sought to determine whether a brief, self-administered, COPD-specific QOL measure, the Seattle Obstructive Lung Disease Questionnaire (SOLDQ), could accurately predict hospitalizations and death.

Design: Prospective cohort study.

Setting: Patients enrolled in the primary care clinics at seven Department of Veterans Affairs (VA) medical centers participating in the Ambulatory Care Quality Improvement Project.

Patients: Of 24,458 patients who completed a health inventory, 5,503 reported having chronic lung disease. The 3,282 patients who completed the baseline SOLDQ were followed for 12 months.

Measurements: Hospitalization and all-cause mortality during the 1-year follow-up period.

Results: During the follow-up period, 601 patients (18.3%) were hospitalized, 141 (4.3%) for COPD exacerbations, and 167 patients (5.1%) died. After adjusting for age, VA hospital site, distance to the VA hospital, employment status, and smoking status, the relative risk of any hospitalization among patients with scores on the emotional, physical, and coping skills scales of the SOLDQ that were in the lowest quartile, when compared to the highest quartile, were 2.0 (95% confidence interval [CI], 1.5 to 2.6), 2.5 (95% CI, 1.9 to 3.4), and 1.9 (95% CI, 1.5 to 2.5), respectively. When hospitalizations were restricted to those specifically for COPD, the odds ratio (OR) for the lowest quartile of physical function was 6.0 (95% CI, 3.1 to 11.5). Similarly, patients in the lowest quartile of physical function also had an increased risk of death (OR, 6.8; 95% CI, 3.3 to 13.8). When adjusted for comorbidity (OR, 0.8; 95% CI, 0.5 to 1.2), long-term steroid use (OR, 2.8; 95% CI, 1.6 to 4.9), and prior hospitalization for COPD (OR, 4.5; 95% CI, 2.2 to 9.2), patients having baseline SOLDQ physical function scores in the lowest quartile had an odds of hospitalization for COPD that was fivefold higher than patients with scores in the highest quartile (OR, 5.0; 95% CI, 2.6 to 9.7).

Conclusions: Lower QOL is a powerful predictor of hospitalization and all-cause mortality. Brief, self-administered instruments such as the SOLDQ may provide an opportunity to identify patients who could benefit from preventive interventions.

Figures in this Article

It is estimated that COPD affects 16 million persons in the United States and is the fourth leading cause of death.1Among Medicare beneficiaries in 1991, there were 132,000 patients hospitalized for COPD exacerbations, with a median length of stay of 7 days.2Per capita expenditures for patients with COPD are approximately 2.4-fold greater than for patients without COPD, and nearly two thirds of the costs are incurred in the hospital.3

Because hospitalizations account for such a major portion of the cost of caring for patients with COPD, it might be helpful to identify those who are at the highest risk of being hospitalized in order to target preventive interventions. Several studies have investigated the risk factors for rehospitalization among patients who have been hospitalized for COPD,45 or hospitalization rates for patients receiving long-term oxygen therapy6or those with severe COPD.7 As these studies have focused on patients with more severe disease, the risk factors associated with hospitalization for the larger number of outpatients seen in general practices may be different.

Patients with COPD who have more frequent exacerbations have worse quality of life (QOL).7 Two studies5,8 using a disease-specific measure of QOL in patients with severe COPD have found that patients with worse scores are more likely to have frequent exacerbations and hospitalizations. These studies found that QOL was an important predictor for hospitalization, and they suggest that disease-specific QOL also may predict exacerbations in patients with less severe COPD.

The goal of the present study was to identify clinical risk factors for hospitalization and death among patients with chronic obstructive lung diseases who are seen in a primary care setting. Specifically, we assessed whether a previously developed disease-specific QOL instrument, the Seattle Obstructive Lung Disease Questionnaire (SOLDQ),9 could be used, in conjunction with other risk factors, to accurately identify patients who are at high risk of hospitalization.

Subjects

The Ambulatory Care Quality Improvement Project (ACQUIP) was a multicenter, randomized, controlled trial that was designed to evaluate quality improvement interventions in a primary care setting. The study population was composed of patients enrolled in general internal medicine clinics (GIMCs) at seven Department of Veterans Affairs (VA) hospitals (ie, Birmingham, AL; Little Rock, AR; San Francisco, CA; West Los Angeles, CA; White River Junction, VT; Richmond, VA; and Seattle, WA).

As part of the ACQUIP study, patients were asked to provide regular assessments of their health and satisfaction with care. This information was linked to data on resource utilization, medication use, and laboratory results that was extracted from the information system of the VA. All patients enrolled in the ACQUIP study were sent a baseline health inventory questionnaire (ie, the Health Checklist), which included a survey of chronic medical conditions, and a general measure of quality-of-life, the medical outcomes study short form-36 (SF-36).10

If a patient reported having any of six chronic diseases (ie, coronary artery disease, chronic lung disease, diabetes, hypertension, depression, and alcohol dependence) on the baseline questionnaire, a disease-specific QOL questionnaire then was sent to that patient (Fig 1 ). Specifically, if a patient reported having “chronic lung disease, emphysema, asthma, or bronchitis,” that patient was sent the SOLDQ. On a chart review of 97 patients during the development and validation of the SOLDQ that was conducted in the same population, medical records were available for 85 subjects. Of these, 83 of 85 patients (97.6%) had COPD or asthma documented in their medical records.9

Baseline cross-sectional data were gathered from 24,458 enrollees, of whom 5,503 (22.5%) had self-reported lung disease, and 3,282 completed the initial SOLDQ. These latter patients, who were outpatients with obstructive lung diseases including both COPD and asthma, comprised the study group.

Predictor Variables

The main predictor variables for the present study were scores on the SOLDQ, a disease-specific QOL questionnaire for patients with COPD.9 The questionnaire is self-administered, written at the fourth-grade level, and requires approximately 5 to 10 min to complete. It consists of 29 items that comprise three health-related QOL dimensions as well as treatment satisfaction. The three QOL dimensions are summarized in the following scales: physical function; emotional function; and coping skills. The satisfaction scale was not considered as a predictor variable, as the primary goal of this study was to assess the predictive value of health-related QOL scales.

The sum of the scores in each dimension is converted to scores ranging from 0 (worst) to 100 (best). No overall score can be generated, and each of the scales is scored separately. The SOLDQ also has been found to be responsive to change in patients whose symptoms have improved or who have recovered from an acute exacerbation. In the initial study describing the development of the SOLDQ, a change of 5 points in an individual scale of the SOLDQ was associated with a clinically detectable change in functional status.

All patients also completed the SF-36, which is a reliable, valid, and responsive measure made up of 36 questions that can be used to compute a physical component summary (PCS) score and a mental component summary (MCS) score. These scores are standardized to the US population (mean score, 50; SD, 10).11

Sociodemographic characteristics of patients were identified from a baseline questionnaire that was sent to all the participants in the ACQUIP study. The questionnaire included questions on race, income, marital status, employment status, education, and smoking status. We also collected information about additional clinical characteristics that may be risk factors for hospitalization or death in patients with COPD. These include age,12 long-term steroid use,6 and prior hospitalization,13 which have been found in previous studies to be associated with hospitalizations and mortality, and therefore are considered to be markers of disease severity in patients with COPD. The variables that were assessed in this study included any hospitalization at a VA facility during the year before enrollment and prior hospitalization for COPD. We defined long-term oral corticosteroid use as either present or absent if a subject had been prescribed oral prednisone for > 90 days of the 180 days preceding enrollment into the study.

Because comorbidity has been found to contribute to mortality in patients with COPD,14we employed two methods of adjustment. First, a commonly used index of comorbidity, the Charlson score,15was calculated for patients who had been hospitalized previously. We employed an adaptation of the Charlson index by Deyo et al,16 which was developed for use with the International Classification of Diseases, ninth revision-clinical modification, (ICD-9-CM) discharge diagnoses. For the 26% of the study participants who had been hospitalized in the 5 years prior to study enrollment, we therefore were able to compute a Charlson score.1516 Second, to assess the comorbidity of those patients who had no prior hospitalization, and for whom no Charlson score could be calculated, we used data from the baseline Health Checklist. Seven of the 24 self-identified chronic conditions on the checklist were associated with an increased risk of hospitalization and mortality in univariate logistic regression (p < 0.10). These conditions were the following: prior myocardial infarction; cancer; congestive heart failure; diabetes; chronic renal insufficiency; history of pneumonia; and stroke.

We performed the analysis both by adjusting for all seven conditions individually, and then by using a count of the number of the conditions that were present. To simplify the final model, we created a dichotomous variable that was positive if two or more conditions were present and negative if two or fewer conditions were present. We found that there was no difference in the point estimates of the other variables in the model, regardless of which method of adjustment for comorbidity we used. We therefore selected the dichotomous variable as the final adjustment for comorbidity.

The likelihood of admission to a VA hospital is influenced by how far away the patient lives from the facility.17To evaluate potential confounding by distance to the VA medical center, we calculated the straight-line distance18 between the patient’s residence and the VA medical center using zip codes to determine longitude and latitude (Geographic Data Technology; Lebanon, NH).

Pulmonary function testing (PFT) has been used as a physiologic marker of disease severity but was not systematically obtained on patients in the study. We obtained the most recent available PFT results for subjects at the Seattle site from the records maintained by the pulmonary function laboratory. Of the 606 patients at the Seattle VA center with lung disease who had returned an SOLDQ, 230 (38%) had undergone PFT. As PFT results were only available for a small percentage of patients, we could not use the variable FEV1 as a separate measure of disease severity in the analysis.

Outcome Variables

The main outcome variables were hospitalization for any reason at the participating VA medical center, hospitalization for a COPD exacerbation at the participating VA medical center, and all-cause mortality during the 1-year period following enrollment in the study. Only the first hospitalization after the completion of the questionnaire was considered in the analysis. Hospital admission and discharge dates, length of hospitalization, and primary and secondary International Classification of Disease, ninth revision (ICD-9) discharge diagnosis codes were obtained from the VA computerized information system at each local site.

Hospitalization for a COPD exacerbation was defined by a primary discharge diagnosis of COPD (ICD-9-CM codes, 490 to 493 or 496) or respiratory infection (ICD-9-CM codes, 460 to 466 or 480 to 487). Admissions to a nursing home or extended care unit were excluded as they were unlikely to reflect an acute COPD exacerbation.

Deaths during the study period were ascertained by the following two mechanisms: (1) weekly interrogation of the VA computerized information system; and (2) the VA Beneficiary Identification and Record Locator Subsystem, which records the dates of death of veterans whose families file for the veteran’s death benefit.19 A combination of these two sources has been found to accurately identify 98.8% of deaths among Medicare-eligible patients.19

Statistical Analysis

To examine possible response bias, we compared the baseline characteristics of patients who completed the SOLDQ with those of patients who had self-identified lung disease but did not complete the SOLDQ, using the Student t test for continuous variables and the χ2 test for categoric variables.

We first constructed logistic models in which QOL scores were modeled as predictor variables without the inclusion of any other clinical risk factors for hospitalization or death. Each scale of the SOLDQ (ie, emotional function, physical function, and coping skills) and the SF-36 (MCS and PCS) was divided into quartiles. The outcome variables for each of these models were any hospitalization, hospitalization for COPD, and mortality.

Potential confounding variables that were entered into the models included demographic characteristics, smoking status, and factors associated with access to the VA such as distance or reported use of non-VA health-care facilities.20 The primary goal of the analysis was to determine the association between QOL and hospitalizations. We therefore started by constructing a model containing only the SOLDQ score and the outcome. We then assessed the demographic variables for potential confounding by adding them to the original model and seeing whether there was a significant change (± 10%) in the coefficient associated with the SOLDQ score.

The only variable that accounted for a change in the estimated coefficients between QOL scales and any hospitalization of > 10% was VA hospital site (ie, the seven sites participating in the ACQUIP study). None of the other potentially confounding variables significantly altered the relationship between QOL and COPD hospitalization or mortality. Distance to the VA, employment status, smoking status, and age were included as they increased precision of the model.

In the final model, we added other measures of disease severity (ie, long-term steroid use, prior hospitalization, and comorbidity) to determine whether the association between SOLDQ and hospitalizations persisted after controlling for these variables. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to estimate the risk of hospitalization for patients with a QOL score in the lowest three quartiles compared to those in the highest (best) quartile using maximum likelihood methods.

After comparing models containing the different scales of the SOLDQ using the likelihood ratio test, a final model containing the physical function scale of the SOLDQ and the additional clinical risk factors for COPD hospitalization was created to estimate the probability of hospitalization in the study population. Discrimination and reliability of the multivariate models were assessed using the c-index and the Hosmer-Lemeshow goodness-of-fit test statistic.21

Among patients with self-identified lung disease, those who had completed the SOLDQ at baseline differed from those who did not complete the questionnaire in several respects (Table 1 ). Compared to nonrespondents, respondents were more likely to be older (mean age, 66 vs 63 years, respectively), to be white (86% vs 78%, respectively), to be married (60% vs 54%, respectively), and to have quit smoking (58.1% vs 51.4%, respectively). Respondents were also less likely to be working full time and to have higher incomes. Approximately 10% of patients in both groups had been hospitalized in the year before study enrollment, including 2% who had been hospitalized for COPD (Table 2 ). Among the 1,435 patients for whom a Charlson score could be calculated, there was a small, though statistically significant, difference in mean scores between respondents and nonrespondents (0.8 vs 0.9, respectively; p = 0.04), while there was no difference in the mean number of self-identified comorbidities (1.3 for both groups). Nonrespondents had lower mean PCS scores compared to respondents (26.6 vs 29.3, respectively) and had a higher 1-year mortality incidence (7.7 vs 5.1, respectively).

The mean (± SD) PCS score for respondents was 29.3 ± 9.9, or two SDs below the mean for the general US population (Table 2). The mean MCS score was 42.6 ± 13.2, or 0.7 SDs below the mean for the general US population. The mean scores on the baseline emotional function, physical function, and coping skills scales of the SOLDQ reflected relatively severe impairment related to COPD (Table 2). In the subset of respondents at the Seattle site for whom pulmonary function test data were available, the mean FEV1 was 1.93 ± 0.77 L (62.4% of predicted).

During the 1-year follow-up period, 601 patients (18.3%) were hospitalized, of whom 68.6% were admitted to acute medicine, cardiology, hematology, or neurology services; 21.7% were admitted to surgical wards; 3.4% were admitted to psychiatric wards; and 6.4% were admitted to intermediate medical, geriatric, or rehabilitation wards. One hundred forty-one patients (4.3%) were hospitalized for a COPD-related illness, of whom 95 had obstructive lung disease (ICD-9 codes, 490 to 496), 45 had pneumonia (ICD-9 codes, 481 to 486), and 1 had acute bronchitis (ICD-9 code, 466). There were 167 deaths (5.1%) during the study period.

Baseline unadjusted SOLDQ physical function scores were associated with both hospitalizations and mortality. Those patients in the lowest quartile (worst scores) of physical function scores were hospitalized more than twice as often for any reason during the subsequent year as patients in the highest quartile (25.5% vs 12.0%, respectively; Fig 2 ), and were hospitalized five times as often for COPD (8.5% vs 1.7%, respectively). Moreover, the mortality rate among patients in the lowest quartile was more than seven times higher (9.3% vs 1.3%, respectively). Similar results were found using the baseline SOLDQ emotional function and coping skills scores (data not shown).

After adjustment for age, site, distance from the VA, employment status, and smoking status, there was an increased risk of a COPD-related hospitalization for patients with SOLDQ physical function scores in the lowest quartile compared to those in the highest quartile (OR, 6.0; 95% CI, 3.1 to 11.5) [Table 3] . There was also an increased risk of any hospitalization (OR, 2.5; 95% CI, 1.9 to 3.4) and death (OR, 6.8; 95% CI, 3.3 to 13.8). Scores on the SOLDQ emotional function and coping skills scales were similarly predictive of hospitalization and death. These findings remained unchanged when events occurring in the first 2 months of follow-up after completion of the SOLDQ were excluded.

Patients with SF-36 PCS scores in the lowest quartile also had an increased risk of hospitalization and death (Table 3). The magnitude of the risk associated with the PCS score compared to the SOLDQ physical function score was lower for COPD hospitalization (OR, 2.5 vs 6.0, respectively) and death (OR, 1.8 vs 6.8, respectively). The SF-36 MCS score was unassociated with the risk of hospitalization for COPD.

In a multivariate logistic regression model that included other clinical variables associated with severe disease, the relationship between baseline SOLDQ scores and hospitalization and mortality rates remained statistically significant. When combined in the same multivariate model, SOLDQ physical function score, age, long-term steroid use, comorbidity, and prior hospitalization were all significantly associated with any hospitalization and with death (Table 4 ). Comorbidity, as assessed by two or more self-reported chronic conditions, was not a risk factor for hospitalization for COPD (OR, 0.8; 95% CI, 0.5 to 1.2).

In other, similarly constructed multivariate models, the SOLDQ emotional function and coping skills scores were also strongly associated with hospitalization and death, although the ORs were lower than for physical function (data not shown). When the emotional function or coping skills scales were added to a model containing physical function alone, there was no additional improvement in predictive accuracy. This was true for models in which either hospitalizations or mortality were the outcomes.

Combining the SOLDQ physical function score with other clinical measures of disease severity in a logistic model provided accurate discrimination with respect to the risk of hospitalization for COPD (c-index, 0.706; Hosmer-Lemeshow test statistic, 7.9; p = 0.4) [Table 5] . For example, a 70-year-old patient who is a long-term user of steroids, was hospitalized for COPD in the past year, and had an SOLDQ physical function score in the lowest quartile has a 44.9% probability of a hospitalization for COPD during the subsequent year (95% CI, 28.3 to 62.7). On the other hand, a 60-year-old patient who does not use steroids, has not been hospitalized for COPD, and has an SOLDQ physical function score in the highest quartile has only a 1.5% probability of hospitalization (95% CI, 1.0 to 2.3).

We found that the SOLDQ, a condition-specific QOL measure, was an independent and important predictor for subsequent hospitalization, hospitalization for COPD, and mortality. Lower scores on all three scales of the SOLDQ were associated with an increasing risk of hospitalization and death. The risk of COPD-related hospitalization is of importance since hospital admissions account for most of the costs of caring for patients with COPD.3 Identifying high-risk patients may allow the development of interventions that would reduce this risk.

Of the three SOLDQ scales, the physical function scale was the best predictor of hospitalization for COPD, indicating that patients with the least physical reserve are at the greatest risk. The coping scale was also a significant predictor and suggests that patients who are the least able to cope with their COPD could potentially benefit from interventions designed to improve their coping skills.

Although both the SOLDQ and the SF-36 were associated with COPD-related hospitalizations, the SOLDQ exhibited much greater power as a predictor than the summary scores of the SF-36 (ie, MCS and PCS). Only subjects with scores in the lowest two quartiles of the PCS score had an increased risk of hospitalization, and the MCS score of the SF-36 was not associated with COPD hospitalizations.

These results are consistent with previous studies of patients with moderate-to-severe COPD. In a study7 of COPD patients with severe disease (mean FEV1, 0.7 L), patients with lower scores on the sickness impact profile, which is a general measure of health-related QOL, had more frequent exacerbations and hospitalizations. Two more recent studies have used a respiratory-specific QOL instrument, the St. George’s Respiratory Questionnaire (SGRQ).,5,8 Among patients who were hospitalized for a COPD exacerbation, lower scores on the SGRQ predicted hospital readmission within the next 12 months, while sex, age, and pulmonary function did not. In another study8 of 61 stable outpatients, those with low SGRQ scores had more frequent exacerbations.

When we combined the baseline SOLDQ physical function score with information from the clinical history, the resulting predictive model was well-calibrated and exhibited good discrimination. Using a logistic model, we could identify groups of patients who had a 1-year risk of hospitalization for COPD within the next year of as low as a 2%, or as high as a 45%. This equation provides different cutoffs that could be applied in different clinical settings to identify groups of patients with COPD who are at high risk for hospitalization.

The results of this study suggest that reliable and valid measures of condition-specific QOL, such as the SOLDQ, can provide important prognostic information about the course of COPD. The SOLDQ is self-administered, is written at a fourth-grade level, and can be completed in 5 to 10 min while a patient waits for an appointment. The other clinical risk factors are also easily obtainable and provide a convenient method with which to identify patients who are at high risk of hospitalization or death. Not only might this information be of interest to clinicians and patients, but it could help to target disease-management interventions such as more intensive primary care follow-up, specialty referral, pulmonary rehabilitation, or social or coping interventions.

We also found that SOLDQ scores were independent predictors for mortality in this population. The most significant risk factors for mortality in COPD from previous trials have been found to be age and FEV1.12 Several studies2223 have suggested, however, that functional status is related to mortality after hospitalization in the elderly. Anthonisen and colleagues23 noted that in COPD patients, the sickness impact profile physical scale also was related to mortality. Similar to predicting hospitalization, prognostic information about mortality may be useful for clinicians and patients.

One of the strengths of this study is that the study participants were outpatients being followed in GIMCs across the United States and, therefore, were more representative of patients who are seen in general clinical practice than those in prior studies. Patients were mailed the questionnaires at home, and therefore these patients are more likely to be clinically stable than if given the questionnaire as inpatients or during clinic visits. Although some patients may have been experiencing an acute exacerbation when completing the questionnaire and then were hospitalized shortly after completing the SOLDQ, the estimates of the risk of hospitalization did not change appreciably, when excluding all patients who had been hospitalized in the first 2 months after completing the questionnaire. Other strengths of this study were that it was population-based with a large sample size and that extensive information was obtained about nonresponders.

There were also several potential limitations to this study. The study population of patients with COPD was determined using a self-administered questionnaire in which patients were asked whether they had chronic lung disease. On a chart review of 97 patients, 97.6% had COPD or asthma in their medical records. Because we could not confirm the diagnosis with PFT, the study population likely included some patients with asthma as well as a small number with nonobstructive lung disease. In addition, although approximately 20% of patients in the overall study population reported having lung disease, there may have been patients with COPD who did not report lung disease on the initial Health Checklist and were not included in the study.

As pulmonary function tests were unavailable for most patients in this study, we were not able to use FEV1 in the final model to predict COPD hospitalization. It is possible that FEV1 also may have been an important predictor of hospitalization and mortality in this population. A recent study24 has suggested that physiologic data such as PFT may measure a different dimension of disease severity than QOL, and further research is necessary to compare the relative contributions of these two sources of information.

Another potential limitation was the incomplete response to mailed questionnaires. The extensive data that we gathered on nonrespondents suggest that there were several differences in demographic and disease-severity measures between the study sample and those patients who were excluded because they did not complete the SOLDQ. These results may not, therefore, be applicable to patients who tend not to return questionnaires. Finally, the patients in the study were male veterans who obtain their primary care at VA hospitals. These factors limit the generalizability of these findings to women and non-VA medical center populations.

In conclusion, this study has shown that QOL in COPD patients, measured using the SOLDQ, was a significant independent risk factor for both hospitalizations and death. Using the SOLDQ physical function scale in a model that included other clinical risk factors for hospitalization, it was possible to identify a group of patients at high risk for hospitalization for a COPD exacerbation, providing an opportunity to identify patients who would potentially benefit from preventive measures. Future research is needed to validate our prediction model in other populations of patients with COPD, as our sample was not large enough to both develop and validate the model.

Abbreviations: ACQUIP = Ambulatory Care Quality Improvement Project; CI = confidence interval; GIMC = general internal medicine clinic; ICD-9 = International Classification of Diseases, ninth revision; ICD-9-CM = International Classification of Diseases, ninth revision-clinical modification; MCS = mental component summary; OR = odds ratio; PCS = physical component summary; PFT = pulmonary function test; QOL = quality of life; SF-36 = medical outcomes study short form-36; SGRQ = St. George’s Respiratory Questionnaire; SOLDQ = Seattle Obstructive Lung Disease Questionnaire; VA = Department of Veterans Affairs

The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.

The research reported here was supported by the Department of Veteran Affairs, Veterans Health Administration, Health Services Research and Development Service grants SDR 96-002 and IIR 99-376.

Figure Jump LinkFigure 1. Identification of eligible subjects with COPD who completed the baseline SOLDQ.Grahic Jump Location
Table Graphic Jump Location
Table 1. Demographic Characteristics of Patients With Self-Identified Lung Disease*
* 

Values given as mean (SD), unless otherwise indicated.

 

Missing (n = 224).

 

Based on a demographic questionnaire that asks whether a patient also obtains care outside the VA system.

§ 

Straight-line distance calculated from patient zip code to VA hospital zip code.

Table Graphic Jump Location
Table 2. Disease Severity and Comorbidity of Patients With Self-Identified Lung Disease*
* 

Values given as % or mean (SD), unless otherwise indicated. NA = not applicable

 

In the 1-year period prior to the index date.

 

Values given as mean score.

Figure Jump LinkFigure 2. Unadjusted hospitalization and mortality rates according to baseline SOLDQ physical function score.Grahic Jump Location
Table Graphic Jump Location
Table 3. One-Year Risk of Hospitalization or Death According to Baseline SOLDQ or SF-36 Scores*
* 

Values adjusted for age, site, distance from VA, employment status, and smoking status.

Table Graphic Jump Location
Table 4. Multivariate Model of Risk of Hospitalization and Mortality Based on SOLDQ Physical Function Score and Clinical History*
* 

Values adjusted for site, distance, employment, and smoking status.

 

OR associated with two or more comorbid diseases.

 

Hospitalization within 1 year prior to index date.

Table Graphic Jump Location
Table 5. Predicted 1-Year Risk of COPD and Identification of High-Risk Subgroups
Benson, V, Marano, M. (1998) Current estimates from the National Health Interview Survey, 1995.Vital Health Stat10,12-13
 
Cydulka, RK, McFadden, ER, Jr, Emerman, CL, et al Patterns of hospitalization in elderly patients with asthma and chronic obstructive pulmonary disease.Am J Respir Crit Care Med1997;156,1807-1812. [PubMed]
 
Grasso, ME, Weller, WE, Shaffer, TJ, et al Capitation, managed care, and chronic obstructive pulmonary disease.Am J Respir Crit Care Med1998;158,133-138. [PubMed]
 
Connors, AF, Jr, Dawson, NV, Thomas, C, et al Outcomes following acute exacerbation of severe chronic obstructive lung disease: the SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments) [published erratum appears in Am J Respir Crit Care Med 1997; 155:386]Am J Respir Crit Care Med1996;154,959-967. [PubMed]
 
Osman, IM, Godden, DJ, Friend, JA, et al Quality of life and hospital re-admission in patients with chronic obstructive pulmonary disease.Thorax1997;52,67-71. [PubMed] [CrossRef]
 
Strom, K Oral corticosteroid treatment during long-term oxygen therapy in chronic obstructive pulmonary disease: a risk factor for hospitalization and mortality in women.Respir Med1998;92,50-56. [PubMed]
 
Traver, GA Measures of symptoms and life quality to predict emergent use of institutional health care resources in chronic obstructive airways disease.Heart Lung1988;17,689-697. [PubMed]
 
Seemungal, TA, Donaldson, GC, Paul, EA, et al Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease.Am J Respir Crit Care Med1998;157,1418-1422. [PubMed]
 
Tu, SP, McDonell, MB, Spertus, JA, et al A new self-administered questionnaire to monitor health-related quality of life in patients with COPD: Ambulatory Care Quality Improvement Project (ACQUIP) Investigators.Chest1997;112,614-622. [PubMed]
 
Ware, JE, Jr, Gandek, B Overview of the SF-36 Health Survey and the International Quality of Life Assessment (IQOLA) Project.J Clin Epidemiol1998;51,903-912. [PubMed]
 
Ware, JE, Jr, Kosinski, M, Bayliss, MS, et al Comparison of methods for the scoring and statistical analysis of SF-36 health profile and summary measures: summary of results from the Medical Outcomes Study.Med Care1995;33,AS264-AS279. [PubMed]
 
Anthonisen, NR Prognosis in chronic obstructive pulmonary disease: results from multicenter clinical trials.Am Rev Respir Dis1989;140,S95-S99. [PubMed]
 
Vestbo, J, Prescott, E, Lange, P, et al Vital prognosis after hospitalization for COPD: a study of a random population sample.Respir Med1998;92,772-776. [PubMed]
 
Incalzi, A, Fuso, L, De Rosa, M, et al Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease.Eur Respir J1997;10,2794-2800. [PubMed]
 
Charlson, ME, Pompei, P, Ales, KL, et al A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.J Chronic Dis1987;40,373-383. [PubMed]
 
Deyo, RA, Cherkin, DC, Ciol, MA Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.J Clin Epidemiol1992;45,613-619. [PubMed]
 
Burgess, JF, Jr, DeFiore, DA The effect of distance to VA facilities on the choice and level of utilization of VA outpatient services.Soc Sci Med1994;39,95-104. [PubMed]
 
Phibbs, CS, Luft, HS Correlation of travel time on roads versus straight line distance.Med Care Res Rev1995;52,532-542. [PubMed]
 
Fleming, C, Fisher, ES, Chang, CH, et al Studying outcomes and hospital utilization in the elderly: the advantages of a merged data base for Medicare and Veterans Affairs hospitals.Med Care1992;30,377-391. [PubMed]
 
Bindman, AB, Grumbach, K, Osmond, D, et al Preventable hospitalizations and access to health care.JAMA1995;274,305-311. [PubMed]
 
Hosmer, DJ, Lemeshow, S. Applied logistic regression. 1989; Wiley-Interscience. New York, NY:.
 
Inouye, SK, Peduzzi, PN, Robison, JT, et al Importance of functional measures in predicting mortality among older hospitalized patients.JAMA1998;279,1187-1193. [PubMed]
 
Anthonisen, NR, Wright, EC, Hodgkin, JE Prognosis in chronic obstructive pulmonary disease.Am Rev Respir Dis1986;133,14-20. [PubMed]
 
Wijnhoven, HA, Kriegsman, DM, Hesselink, AE, et al Determinants of different dimensions of disease severity in asthma and COPD: pulmonary function and health-related quality of life.Chest2001;119,1034-1042. [PubMed]
 

Figures

Figure Jump LinkFigure 1. Identification of eligible subjects with COPD who completed the baseline SOLDQ.Grahic Jump Location
Figure Jump LinkFigure 2. Unadjusted hospitalization and mortality rates according to baseline SOLDQ physical function score.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1. Demographic Characteristics of Patients With Self-Identified Lung Disease*
* 

Values given as mean (SD), unless otherwise indicated.

 

Missing (n = 224).

 

Based on a demographic questionnaire that asks whether a patient also obtains care outside the VA system.

§ 

Straight-line distance calculated from patient zip code to VA hospital zip code.

Table Graphic Jump Location
Table 2. Disease Severity and Comorbidity of Patients With Self-Identified Lung Disease*
* 

Values given as % or mean (SD), unless otherwise indicated. NA = not applicable

 

In the 1-year period prior to the index date.

 

Values given as mean score.

Table Graphic Jump Location
Table 3. One-Year Risk of Hospitalization or Death According to Baseline SOLDQ or SF-36 Scores*
* 

Values adjusted for age, site, distance from VA, employment status, and smoking status.

Table Graphic Jump Location
Table 4. Multivariate Model of Risk of Hospitalization and Mortality Based on SOLDQ Physical Function Score and Clinical History*
* 

Values adjusted for site, distance, employment, and smoking status.

 

OR associated with two or more comorbid diseases.

 

Hospitalization within 1 year prior to index date.

Table Graphic Jump Location
Table 5. Predicted 1-Year Risk of COPD and Identification of High-Risk Subgroups

References

Benson, V, Marano, M. (1998) Current estimates from the National Health Interview Survey, 1995.Vital Health Stat10,12-13
 
Cydulka, RK, McFadden, ER, Jr, Emerman, CL, et al Patterns of hospitalization in elderly patients with asthma and chronic obstructive pulmonary disease.Am J Respir Crit Care Med1997;156,1807-1812. [PubMed]
 
Grasso, ME, Weller, WE, Shaffer, TJ, et al Capitation, managed care, and chronic obstructive pulmonary disease.Am J Respir Crit Care Med1998;158,133-138. [PubMed]
 
Connors, AF, Jr, Dawson, NV, Thomas, C, et al Outcomes following acute exacerbation of severe chronic obstructive lung disease: the SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments) [published erratum appears in Am J Respir Crit Care Med 1997; 155:386]Am J Respir Crit Care Med1996;154,959-967. [PubMed]
 
Osman, IM, Godden, DJ, Friend, JA, et al Quality of life and hospital re-admission in patients with chronic obstructive pulmonary disease.Thorax1997;52,67-71. [PubMed] [CrossRef]
 
Strom, K Oral corticosteroid treatment during long-term oxygen therapy in chronic obstructive pulmonary disease: a risk factor for hospitalization and mortality in women.Respir Med1998;92,50-56. [PubMed]
 
Traver, GA Measures of symptoms and life quality to predict emergent use of institutional health care resources in chronic obstructive airways disease.Heart Lung1988;17,689-697. [PubMed]
 
Seemungal, TA, Donaldson, GC, Paul, EA, et al Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease.Am J Respir Crit Care Med1998;157,1418-1422. [PubMed]
 
Tu, SP, McDonell, MB, Spertus, JA, et al A new self-administered questionnaire to monitor health-related quality of life in patients with COPD: Ambulatory Care Quality Improvement Project (ACQUIP) Investigators.Chest1997;112,614-622. [PubMed]
 
Ware, JE, Jr, Gandek, B Overview of the SF-36 Health Survey and the International Quality of Life Assessment (IQOLA) Project.J Clin Epidemiol1998;51,903-912. [PubMed]
 
Ware, JE, Jr, Kosinski, M, Bayliss, MS, et al Comparison of methods for the scoring and statistical analysis of SF-36 health profile and summary measures: summary of results from the Medical Outcomes Study.Med Care1995;33,AS264-AS279. [PubMed]
 
Anthonisen, NR Prognosis in chronic obstructive pulmonary disease: results from multicenter clinical trials.Am Rev Respir Dis1989;140,S95-S99. [PubMed]
 
Vestbo, J, Prescott, E, Lange, P, et al Vital prognosis after hospitalization for COPD: a study of a random population sample.Respir Med1998;92,772-776. [PubMed]
 
Incalzi, A, Fuso, L, De Rosa, M, et al Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease.Eur Respir J1997;10,2794-2800. [PubMed]
 
Charlson, ME, Pompei, P, Ales, KL, et al A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.J Chronic Dis1987;40,373-383. [PubMed]
 
Deyo, RA, Cherkin, DC, Ciol, MA Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.J Clin Epidemiol1992;45,613-619. [PubMed]
 
Burgess, JF, Jr, DeFiore, DA The effect of distance to VA facilities on the choice and level of utilization of VA outpatient services.Soc Sci Med1994;39,95-104. [PubMed]
 
Phibbs, CS, Luft, HS Correlation of travel time on roads versus straight line distance.Med Care Res Rev1995;52,532-542. [PubMed]
 
Fleming, C, Fisher, ES, Chang, CH, et al Studying outcomes and hospital utilization in the elderly: the advantages of a merged data base for Medicare and Veterans Affairs hospitals.Med Care1992;30,377-391. [PubMed]
 
Bindman, AB, Grumbach, K, Osmond, D, et al Preventable hospitalizations and access to health care.JAMA1995;274,305-311. [PubMed]
 
Hosmer, DJ, Lemeshow, S. Applied logistic regression. 1989; Wiley-Interscience. New York, NY:.
 
Inouye, SK, Peduzzi, PN, Robison, JT, et al Importance of functional measures in predicting mortality among older hospitalized patients.JAMA1998;279,1187-1193. [PubMed]
 
Anthonisen, NR, Wright, EC, Hodgkin, JE Prognosis in chronic obstructive pulmonary disease.Am Rev Respir Dis1986;133,14-20. [PubMed]
 
Wijnhoven, HA, Kriegsman, DM, Hesselink, AE, et al Determinants of different dimensions of disease severity in asthma and COPD: pulmonary function and health-related quality of life.Chest2001;119,1034-1042. [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543