Abstract
Objective: As ALS progresses, extensive supportive care is required, including multidisciplinary outpatient care and hospitalization. The authors studied the causes, health care utilization, and outcomes for hospitalized patients with ALS.
Methods: With use of the 1996 Nationwide Inpatient Sample, an administrative database representing 20% of U.S. hospitals, 1,600 hospitalizations in patients with ALS were identified and compared with 5,364,728 non-ALS hospitalizations.
Results: The most common concurrent diagnoses in patients with ALS were dehydration and malnutrition (574 patients, 36%), pneumonia (507 patients, 32%), and respiratory failure (398 patients, 25%). Only 38% of patients with ALS were discharged to home without home health care compared with 73% of patients with non-ALS. Fifteen percent of patients with ALS died in the hospital compared with 3% of non-ALS patients. The average length of hospital stay and charges were greater for patients with ALS than for non-ALS patients (8.4 days and $19,810 for ALS patients and 5.4 days and $11,924 for non-ALS patients). Mortality was significantly associated with emergency room admission (versus nonemergency admission; OR = 1.60), increasing age (per year; OR = 1.03), respiratory failure (OR = 3.37), and pneumonia (OR = 2.02) (p < 0.01 for all comparisons).
Conclusions: Patients with ALS have lengthy and costly hospital admissions, a high in-hospital mortality rate, and few routine discharges. Recognition of the issues that precipitate hospitalization may allow development of preventive strategies.
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the March 27 issue to find the title link for this article.
The median survival for patients with ALS from time of diagnosis is 1 to 3 years.1 The worldwide incidence of ALS is estimated to be 0.6 to 2.6 cases/100,000 people annually.2 As patients with ALS become disabled, they receive supportive care from advanced practice nurses, physical therapists, occupational therapists, respiratory therapists, and speech pathologists and require medical equipment such as automated wheelchairs, computerized communication devices, feeding tubes, and ventilatory equipment.
Recent estimates of the cost of ALS care have been based primarily on patient report and expert opinion. These estimates have focused on the costs of diagnosis, supportive equipment, home ventilation, and outpatient nursing care.3-6⇓⇓⇓ There are no cost estimates for in-hospital ALS care, nor are there any systematic evaluations of hospital outcomes in this population.
We studied the costs and outcomes of hospitalization for patients with ALS in the U.S.A.
Methods.
Study design and population.
This is a cross-sectional study of utilization and expenditures for ALS patients hospitalized in 1996. Eligible patients for analysis were ≥18 years old and had a primary or secondary discharge diagnosis of ALS, using the International Classification of Diseases, 9th rev., code 335.20.7 To provide a concurrent frame of reference to interpret the findings in patients with ALS, we compared them with all adult non-ALS patients who were hospitalized during 1996.
Database.
The 1996 Nationwide Inpatient Sample (NIS) was used for this study. The NIS was developed as part of the Healthcare Cost and Utilization Project sponsored by the Agency for Healthcare Research and Quality. It consists of patient-level information about hospital admissions from a stratified sample of U.S. hospitals, designed to approximate 20% of all admissions during 1996. It contains information on nonfederal, short-term hospitals, as defined by the American Hospital Association, from 19 states.
Predictor and outcome variables.
Demographic variables assessed included patient age, race, gender, and income. Structure and process of care factors included admission source, hospital type (teaching versus nonteaching), hospital location (urban versus nonurban), and procedures performed. Outcomes were disposition (routine discharge, short-term hospital, skilled nursing, intermediate care, home health care, and death), length of stay, and total charges. Routine discharge is defined as discharge to home without home health care.
To describe the most common comorbid illnesses and procedures, the most frequent diagnoses and procedures were tabulated. The primary diagnosis is identified at the end of the episode of health care as the condition primarily responsible for the patient’s need for treatment or investigation.8 Principal procedures are performed for definitive therapy or to treat a complication and are most closely related to the primary diagnosis.
Because patients with ALS die from pneumonia, respiratory failure, malnutrition, and dehydration,9,10⇓ we identified diagnostic codes associated with these clinical categories.
Because respiratory illnesses and feeding difficulties were the most common expected diagnoses, we also examined procedures that were likely to be performed for pulmonary disorders (pneumonia and respiratory failure) or feeding problems (dehydration and malnutrition). As the cost of neurodiagnostic procedures has been emphasized in the previous literature,6 we also described the frequency of use of these procedures. The procedures to be included in these three categories were determined by review of all procedures coded for patients with ALS and categorizing them as pulmonary, gastrointestinal, neurodiagnostic, or other. Additional details of the methods are available at www.neurology.org.
Analyses.
Descriptive results were tabulated with proportions and means, as appropriate. Patients with ALS were compared with those without ALS. χ2 Testing was used to compare frequencies of categorical variables. When the independent variable had more than two ordered categories, a χ2 test for trend was used. Analysis of variance was used to compare the association of categorical predictor variables with continuous outcome variables. Significance for all tests was set at p < 0.01. Mean values are presented as means ± SD. Statistical analyses were performed using SAS version 7.0 (SAS Institute, Cary, NC) and STATA version 6 (College Station, TX).
Results.
In this 20% sample of U.S. hospitalizations, there were 1,600 ALS and 5,364,728 non-ALS admissions. Patients with ALS were older than other hospitalized patients, more likely to be men and Caucasian and to have higher incomes (p < 0.001 for each comparison) ( table 1).
Patient characteristics and hospital information
More than half of the patients with ALS were admitted through the emergency room, whereas relatively few patients were admitted from other hospitals or other facilities (see table 1). One-third of ALS admissions were to teaching hospitals, and this was not statistically different from that for non-ALS patients (see table 1). Most patients with ALS were admitted to urban hospitals, and Medicare was the primary payer for 63.7%.
Outcomes.
Fifteen percent of ALS patient hospitalizations resulted in death compared with 3% of non-ALS patients ( table 2). Of the patients with ALS who survived hospital admission, 54.6% required further care (short-term hospital, skilled nursing, intermediate care, or home health care) after discharge, and only 45.2% of surviving patients with ALS were discharged routinely to home. In comparison, most surviving non-ALS hospitalized patients were discharged routinely to home.
Outcomes of hospitalized patients
The mean length of stay for patients with ALS was significantly greater than for non-ALS patients, as were mean charges per hospitalization (see table 2). Total charges for patients with ALS ranged from $445 to $884,369 per hospitalization.
Comorbid illnesses and conditions.
The most common primary diagnoses in patients with ALS, other than ALS, were acute respiratory failure, aspiration pneumonitis, and pneumonia with no organism specified ( table 3). Acute respiratory failure was also the most common secondary diagnosis. As anticipated, admissions frequently included conditions that could be grouped within certain diagnostic categories, including dehydration and malnutrition (n = 574; 35.9%), pneumonia (n = 507; 31.7%), and respiratory failure (n = 398; 24.9%). Of the 245 patients with ALS who died, the most common primary diagnoses other than ALS were acute respiratory failure (n = 59; 24.0%), aspiration pneumonitis (n = 33; 13.5%), and pneumonia (n = 20; 8.2%).
Most frequent diagnoses and procedures in hospitalized patients with ALS
In-hospital procedures.
The most frequent procedures are listed in table 3. Percutaneous endoscopic gastrostomy, mechanical ventilation, and tracheostomy were the three most commonly coded primary procedures. There were 414 patients (25.9%) who had pulmonary procedures, 353 patients who underwent gastrointestinal procedures (22.1%), and 87 (5.4%) with neurodiagnostic procedures.
Predictors of hospitalization outcomes.
Charges and length of stay.
Bivariate analysis showed no significant association between age, race, or income with total charges (data not shown). However, charges were greater for men than women ($22,697 versus $15,933; p < 0.001) and those with greater comorbidity (Deyo–Charlson ≥ 2: $23,288 versus Deyo–Charlson = 0: $17,281; p < 0.001). Male gender and increased comorbidity remained significant in multivariate analyses ( table 4). Older age and greater comorbidity were significantly associated with longer length of stay, whereas gender, race, and income were not. In multivariate analysis, increased comorbidity was a significant predictor of length of stay beyond 5 days (see table 4), and increasing age showed a trend toward longer length of stay (OR = 1.12, p = 0.03).
Multivariate analyses to predict outcomes of hospitalization
Mortality.
In-hospital mortality increased significantly with advancing age but was not associated with gender, race, income, comorbidity score, hospital type, or hospital location. Of patients under 40 years of age, only 4.3% died, whereas 18.3% of those 80 years or older died. Nonemergency admissions resulted in death 10.7% of the time, and admissions through the emergency room resulted in death 19.3% of the time (p < 0.001). Mortality was significantly higher when mechanical ventilation was the primary procedure (21.2%) than when it was not (12.9%). Patients with gastrostomy tube insertion as the principal procedure had a mortality rate of 7.9% compared with 15.3% when gastrostomy tube insertion was not the principal procedure. When respiratory failure or pneumonia was the primary diagnosis, patients had significantly higher mortality than patients without those diagnoses (respiratory failure: 32.8 versus 14.1%; pneumonia 21.3 versus 14.1%). Malnutrition and dehydration were not significantly associated with mortality.
Emergency admission, increasing age, respiratory failure, and pneumonia remained significantly associated with in-hospital mortality after adjusting for comorbidity, race, gender, income level, malnutrition, gastrostomy, and mechanical ventilation (see table 4).
Discussion.
We used a large, nationwide administrative database to characterize health care utilization by hospitalized patients with ALS. Pulmonary problems such as pneumonia and respiratory failure were common concomitant diagnoses, as were complications associated with malnutrition and dehydration. In-hospital mortality was five times higher than for non-ALS patients, and total charges were 1.7 times higher. If the 1,600 hospitalizations for ALS from this 20% sample represented by the NIS are extrapolated to all hospital admissions in the United States, there were ≈8,000 hospitalizations for ALS in 1996. Based on a midpoint ALS prevalence estimate of 5 cases per 100,000 people11 and the US population from 1996 of 265,190,000,12 there was ≈1 admission for every 2 to 3 patients with ALS in 1996, of which 15% resulted in in-hospital death.
Two thirds of hospital admissions were to nonteaching hospitals. As more is learned about ALS and new therapies arise, ALS care is becoming more specialized. There is a great deal of supportive care that is unique to ALS, and several authors have suggested that ALS care should be provided by a specialized multidisciplinary “team approach.”13-16⇓⇓⇓ Virtually all ALS centers designated by the Muscular Dystrophy Association are located at academic medical centers. A recent publication using data from the ALS Patient CARE Database, a national prospective survey of patients with ALS and providers, reported that 91% of patients enrolled in the database were enrolled by academic centers.17 Our findings support the possibility that many patients with ALS receive outpatient care from specialty centers but in emergency situations are admitted to community hospitals. This pattern may be due to the closer proximity of community hospitals or may reflect the fact that many patients seek second opinions from academic centers but receive their longitudinal care from local practitioners.
Our findings should be representative of patients with ALS throughout the United States. Questions have arisen about the accuracy of information contained in administrative databases, and numerous studies have addressed these issues.18-20⇓⇓ A study that compared claims data with a prospectively collected clinical database of cardiac catheterization patients19 showed that agreement was high between the two sources for certain diseases including diabetes mellitus and peripheral vascular disease but was lower for other conditions such as angina and tobacco use. Medicare data have been shown to be reliable for estimating incidence rates of cancer but are not as good for estimating cancer resection rates.20 These studies suggest that administrative databases are accurate for many conditions but not all. ALS has well-defined criteria21 and, because of the gravity of the illness, is unlikely to be overdiagnosed. Studies of administrative databases suggest that if they are inaccurate, it is likely that they underestimate the incidence of certain diagnoses. If early cases of ALS have not yet been diagnosed, our findings may underestimate the number of hospitalizations for ALS. It is important to emphasize that the NIS database does not contain unique patient identifiers, and therefore some of the hospitalizations represent repeat admissions for the same patient. This limits our ability to make patient-level predictions about future events, as there is no way to follow individual patients longitudinally.
It is unlikely that prolonged hospitalization maximizes patient quality of life, especially if the hospitalization terminates in death. In light of our findings, it is clear that clinicians need to discuss end-of-life decisions early and throughout the course of ALS. A recent study of patients with ALS revealed that patients were able to express their preferences for various interventions, but that these preferences change over time, necessitating clinical education efforts throughout the disease course.22 Whereas hospital admission can be useful to facilitate procedures such as placement of gastrostomy tubes, many illnesses including pneumonia might be treated as effectively as outpatients. Additionally, hospice is a valuable resource for patients with this devastating disease. Hospice and home care have been shown to improve patient satisfaction while decreasing hospital costs,23,24⇓ and patients with ALS can be treated at home even when there are multiple symptoms and the need for medications.25,26⇓ A recent cohort study of patients with ALS found that 61% of patients died at home, 27% died in a hospital or skilled nursing facility, and 4% died in a hospice. Only 46% of patients were helped by hospice in the terminal phase of the disease.17 These findings are consistent with our data showing that a relatively high percentage of patients with ALS die in hospitals and suggest that end-of-life planning is not currently being optimized. Another important component of patient education is discussing the need for postmortem examination. This is an endeavor that may help determine the pathophysiology of ALS.
Although it is not surprising that patients with a rapidly fatal disease die more frequently than other hospitalized adults, it is of concern that their hospitalizations are frequent, lengthy, and costly. Future studies should examine costs and outcomes of all care provided for patients with ALS. By better quantifying health care expenditures and patient outcomes in ALS, patients, physicians, and health care administrators will be able to make properly informed decisions about many aspects of care. Until more definitive therapies for ALS are available, increased efforts need to be made toward maximizing quality of life by improving patient education, communication, and discussions about end-of-life care throughout the course of the disease.
Acknowledgments
Data for this project were provided by the Healthcare Cost and Utilization Project, 1988 to 1996.
Acknowledgment
The authors thank Drs. Kevin Frick and Jeffrey Rothstein for their thoughtful review of an earlier version of this manuscript.
- Received August 28, 2000.
- Accepted December 2, 2000.
References
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.