INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthma and chronic obstructive pulmonary disease (COPD; e.g., chronic bronchitis and emphysema) are very common illnesses with a great deal of morbidity (1). It is estimated that these two conditions may affect as many as 20 million patients in the U.S. (6), with a total yearly cost in excess of 6.2 billion dollars (6). Neither disorder is typically considered as a major affliction of the elderly but, in point of fact, a significant proportion of the geriatric population (i.e., persons 65 yr or older) are at risk. For example, asthma may involve 7-10% of this population and the prevalence of COPD is thought to range from 13-35% in older smokers (6). Thus, these illnesses occur frequently enough to cause a fair amount of morbidity.

Even though the elderly are the most rapidly growing segment of the population, little is known about the impact of asthma and COPD on health care utilization, particularly inpatient services (10). The present study was designed to fill that void by establishing a temporal profile of the number of hospitalizations, duration of stay, disposition patterns, and financial consequences in patients over 65 yr with these conditions.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The Health Care Financing Administration (HCFA) maintains a computerized database of all hospital discharges for persons covered by Part A of the Medicare program. This database encompasses 96% of all of the people over the age of 65 yr in the United States and is derived from a standardized billing form (HCFA FORM 1450, UB-82). This information is combined annually into a file that contains a principle diagnosis and up to four additional illness coded according to the international Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) for each hospitalization (12). The principle diagnosis is listed first on the UB-82 forms that hospitals submit to HCFA for reimbursement. Other diagnoses typically follow in descending order of significance to the hospitalization. The file also includes the number and types of medical procedures performed, incurred charges, demographic data, and hospital admission/discharge dates.

The entire sample for the year 1984 through 1991 involving beneficiaries  65 yr were searched for the principle diagnoses of asthma (ICD 493.0, 493.1, 493.9) and COPD (ICD 490, 491, 492, 496). The study cohort for the present investigation was created by determining each patient's first admission each year for an exacerbation of either disease from January 1, 1984 through December 31, 1991. Patients were identified by their social security number so as not to be counted more than once per year for each diagnosis. We limited our analysis of length of stay and outcome of hospitalization to only the first hospitalization for each disease each year in order to accurately collect demographic information about our patient population and to avoid skewing data toward those patients with multiple hospitalizations. If a patient was admitted for both asthma and COPD exacerbation in the same year, he would be recorded once in that year's asthma cohort and once in that year's COPD cohort. A separate additional analysis was then performed on the 1984 study cohort that included all repeat hospitalizations in each given year in order to calculate rehospitalization rates for this cohort. In order to track rehospitalization patterns over the 8 yr of the investigation, subsequent hospitalizations of only the 1984 cohort were followed through 1991 by tracking the patients' social security numbers. Patients not hospitalized in the year 1984 were not tracked for previous or subsequent admissions. Codes for cystic fibrosis, bronchiectasis, and other forms of obstructive lung disease were not included. The starting date of 1984 was chosen because it represented the first year in which all hospitals in the United States participated in a prospective payment system based on diagnosis-related groups (DRG) (13). Eight years were thought to be sufficient interval to detect any trends that may have occurred over time.

Because the HCFA data represents over 95% of the population studied, statistical analysis of the data would not provide any additional information. All observed differences are real, rather than inferred.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The demographics for the groups with asthma and COPD at the time points of interests are displayed in Table 1. The total hospitalized cohort with both forms of obstructive lung disease was 219,591 people in 1984. By 1991, the number had fallen to 199,732. In 1984, there were 56,692 patients hospitalized with an exacerbation of asthma as a primary diagnosis and 162,899 with COPD. In 1991, the numbers were 67,758 and 131,974, respectively. In 1984, roughly two-thirds of the patients in the asthma population were female and 87% were white. In the COPD group, males predominated but the racial distribution was similar to the asthmatics. The average age of both populations was 74 ± 7 and 75 ± 7 (mean ± SD), respectively (range 65-105). In the 1991 cohort, the male: female distribution in the COPD category had equalized. There were no other major differences between 1984 and 1991. 

The age distribution of the patients in each population is contained in Figure 1. Approximately 80% of the asthmatics ranged between 65 and 79 yr (Figure 1A). As expected, there were fewer subjects in each cell with advancing age. There were no major differences in the pattern of distribution of this variable between 1984 and 1991. In the COPD group about 70% of the patients were between 65 and 79 yr. Again, there were no changes between years of observation.

View larger version (15K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 1.   Age distribution of asthma cohorts in 1984 and 1991 (A). Age distribution of COPD cohorts in 1984 and 1991 (B).

Figure 2 shows the median duration of hospitalization as a function of age at the beginning and end of our study. In 1984, the median time spent in hospital for asthma and COPD was 6 d (interquartile span = 4-9 d) and 7 d (interquartile span = 4-11 d), respectively. These figures are not static and the length of stay increased as the patients grew older irrespective of the form of airflow limitation. The increase in length of stay was not linearly related to age, as we expected. Note that the very old (85 yr or older) spent a median of 7 d in the hospital for both asthma and COPD exacerbation (interquartile span = 4-10 d for asthma, 4-11 d for COPD). This number is approximately 1 d greater than their younger counterparts independent of diagnosis. Most patients with COPD remained hospitalized for 1 d longer than did patients with asthma in 1984, but this difference disappeared by 1991. Over time, the median duration of hospitalization in patients of all ages with COPD except those older than 90 yrwas shortened by a full day, while the duration of hospitalization for patients with asthma remained unchanged in most age groups.

View larger version (31K): [in this window] [in a new window]   Figure 2.   Median length of hospital stay as a function of age.

There were significant regional variations in length of stay. Patients living in the northwest remained in hospital the least time (1984 asthma: range 3.8-5.1 d; COPD: range 4.6-5.6 d). In contrast, those in the northeast stayed the longest (1984 asthma: range 6.4-7.5 d; range COPD 7.8-9.1 d). In 1991, the same pattern held even though the overall length of stay had shortened.

The outcomes of hospitalization as a factor of age for the cohorts at the beginning (1984) and end of our study (1991) are presented in Figure 3. The outcomes of hospitalization from 1984 to 1991 related to diagnosis is depicted in Figure 4. In 1984, the majority of the younger geriatric patients were discharged home. The very old typically went to other care facilities, home with home health care, or died. With respect to the specific diseases, fewer individuals with COPD went home initially and more died. Once again, age was a negative interactive factor in both groups. As the availability of alternative care options increased over time, fewer patients were sent home directly from the hospital, and, by 1991, the use of convalescent and nursing homes and/or home health care had more than doubled.

View larger version (32K): [in this window] [in a new window]   Figure 3.   Outcome of hospitalization related to age for 1984 and 1991 cohorts. Note that there is variation in the Y axes between the graphs.

View larger version (24K): [in this window] [in a new window]   Figure 4.   Outcome of hospitalization related to diagnosis from 1984-1991. Note that there is variation in the Y axes between the graphs.

The death rates in the asthmatics ranged from 0.7% in the 65 to 69 age bracket to 4.2% in the patients older than 90 yr. The percentages for subjects with COPD who died were roughly twice those of the asthmatics of all ages. There was little change in this pattern over the 7 yr of observation.

Hospital readmission was common in both 1984 cohorts. Just 9% of the asthmatics were not readmitted during the 8-yr observation period. Slightly less than half of this group were hospitalized five or more times, and a few patients required as many as 30 admissions. The asthmatics spent a total of 275,458 d as inpatients, 11% of which were in intensive care units (ICU). Readmission statistics for the patients with COPD were similar. Only 14% of patients with COPD did not require rehospitalization through 1991. Forty-eight percent were admitted five or more times and 14% only once more after entry into the 1984 cohort. In total, the 1984 COPD cohort spent over 1 million d in hospital, or more than three times the number of days, than did the 1984 asthmatic cohort of this time. Patients in the 1984 COPD cohort spent 143,015 d (13% of total hospital days) in the ICU, whereas patients in the 1984 asthma cohort spent 41,645 d (11% of total hospital days) in the ICU. Considerable overlap between the principle admitting diagnoses of asthma and COPD was noted when tracking rehospitalization patterns of both 1984 cohorts (Figure 5).

View larger version (15K): [in this window] [in a new window]   View larger version (10K): [in this window] [in a new window]   Figure 5.   Overlap of diagnosis between asthma and COPD during rehospitalization of the 1984 asthma cohort (A). Overlap of diagnosis between asthma and COPD during rehospitalization of the 1984 COPD Cohort (B) 1984-O indicates first hospitalization, 1984-R indicates rehospitalization in 1984.

The charges associated with asthma and COPD per admission are shown in Figure 6. The average charge in 1984 for asthma was $3,582. The corresponding value for COPD was $4,228. Over time, charges increased in a near linear fashion. The patients with COPD were uniformly assessed a higher rate. From 1984 to 1991, the money requested from third party payers doubled and for the last year of observation averaged $7,694 and $8,876 for asthma and COPD, respectively.

View larger version (19K): [in this window] [in a new window]   Figure 6.   Hospitalization charges for asthma and COPD from 1984- 1991.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of the present study demonstrate that asthma and COPD in the elderly have an enormous impact on the utilization of health care resources. Census figures for 1984 indicate that there were 25.5 million people 65 yr or older living in the United States. Of these, perhaps 7-10% carry the diagnosis of asthma and 13-35% have COPD (1, 6, 13). Only 2-3% of the individuals in both groups require inpatient care, yet together asthma and COPD account for 1.4 million hospital days, of which 10-12% are in intensive care units. The net charge to Medicare and other third party payers exceeded 6.2 billion dollars or approximately 1.5% of the 1984 gross national product. Over the 8 yr of observation, the number of patients hospitalized with asthma and COPD exacerbation fell but the cost of treatment rose. We presume the decline in hospitalization reflects health care trends favoring outpatient therapy. By 1991, 8.8 billion was being spent on health care for patients with these disorders (6, 7). The 80% rate of increase in health care expenditures greatly exceeded the 32.7% rise in consumer price index from 1984 through 1991 (14).

Consumption of health care resources varies as a function of the age of the patient and their locale, as well as the presence of disease specific issues. The length of admission increased with advancing age. Furthermore, as the age of the subjects rose, they were less likely to be discharged to their homes, and were more frequently sent to another care facility. This trend has been seen with other diseases such as stroke, pneumonia, heart failure, and hip replacements, for which (15) referrals for health care after discharge has quadrupled (15). It is unclear if the above phenomenon is due to incomplete resolution of the initial insult, the presence of complicating comorbid conditions, or represents a response to a general decline in overall health. Unfortunately, our database does not contain the information to address any of these possibilities.

Psychosocial issues associated with the loss of the ability for independent living may also be important. Here, too, we are limited by our data. The information that we possess does not list the types of places in which the patients resided before admission so it is not clear whether the elderly with asthma and COPD are less able to function independently posthospitalization as prior to it, or whether they are more likely to live in a care facility as they age. Irrespective of the reasons, the net affect of the above trend is to increase health care utilization. Our data concur with those of Skobeloff and associates who reported the typical length of stay for older asthmatics to range from 6.54 to 7.04 d (11). These authors also noted the duration of hospitalization increased in proportion to the patients age, but they did not study people over 70. Mushlin found that the mean length of stay for an exacerbation of COPD lay between 6.9-8.7 d (23).

Local health practices also have a profound influence on utilization. As in our data, Morris and Munasinghe reported that hospital admission rates in the elderly show marked regional fluctuations, particularly in the southeast and northern plains states (12). They correlated high admission rates with local indicators of socioeconomic status, availability of medical resources, occupational lung disease rates, and smoking. In our study, hospital stays in the southeast were in the third quartile (from the top), while those in the northern plains tended to be shorter than average. Such geographic differences are frequently explained by the fact that changes in health care and health care insurance plans tend to be initiated in the west with the trends in the rest of the country lagging several years behind.

The morbidity and mortality of the diseases under study are also widely divergent. Ninety-four percent of the patients with COPD and 98.8% of the asthmatics survived hospitalization in 1984. These data are similar to those in other smaller studies (13, 18). The mortality figure shifted somewhat in 1991 to 94.4% and 98.1% respectively, presumably, because of fluctuation in the application of the diagnostic criteria (e.g., the number of asthma asthmatics rose with time while the number of people with COPD fell). Patients with COPD, however, were uniformly hospitalized for longer periods and their confinements were more costly than their asthmatic counterparts. In hospital death rates were also 4.6 times higher in those with COPD. Here again, age had a detrimental affect. The in-hospital death rate increased six-fold for asthmatics from age 65 to 90 and 2.5-fold for patients with COPD over the same range.

The reason for the marked differences in morbidity and mortality in COPD are undoubtedly reflective in the pathology of the two conditions. The airway obstruction in asthma is essentially reversible and, therefore, the patient would be expected to recover pulmonary function fully or nearly so. In COPD, however, the inflammatory process in the airway results in irreversible changes as they slowly progress with time (18). Eventually, respiratory failure with or without cor pulmonale ensues. Once this happens, the patient's requirement for continuous monitored care increases with its associated hospitalization costs.

One major limitation of the present investigation is that we were dependent on the accuracy of diagnosis coding of ICD-9. We have no way to retrospectively determine whether the patients with asthma and COPD were correctly categorized. The fact that the ratio between admissions for asthma and COPD changed from 1:2 in 1984 to 1:3 in 1991 may reflect the difficulties encountered when trying to differentiate between asthma and COPD in the elderly (27). The diagnostic uncertainties of differentiating between these two disease processes in the elderly is further demonstrated by the overlap in admission diagnoses on rehospitalization in both 1984 cohorts. Furthermore, we are unable to determine how great the diagnostic bias of physicians toward a diagnosis of COPD in older men changed over the years of the study (28). Fisher and colleagues found that sensitivity of claims data from many hospitals ranged from 58-97% for various diagnoses (26). Specificity, however, was more than 99% (23). Thus, the numbers reported in this study are an underestimation or conservative estimate.

The utilization of health care resources by elderly patients with asthma and COPD is immense, both during hospitalization and after discharge. As the elderly age, they require ever increasing quantities of these resources, as demonstrated by longer periods of hospitalization and a greater number of discharges to skilled care facilities. Furthermore, repeated admissions for exacerbation is the rule, rather than the exception. New strategies toward more efficient care of elderly patients with asthma and COPD must be developed as funding for health care resources for the elderly become increasingly scarce.