|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Time Spent in Hospital in the Last Six Months of Life in Patients Who Died of Cancer in OntarioByFrom the Division of Cancer Care and Epidemiology, Queen’s Cancer Research Institute, Queen’s University and Kingston Regional Cancer Center, Kingston, Ontario, Canada. Address reprint requests to William J. Mackillop, MBChB, Kingston General Hospital, Apps Level 4, Kingston, Ontario K7L 2V7, Canada; email: william.mackillop{at}krcc.on.ca
PURPOSE: To describe hospital bed utilization in the final 6 months of life in patients dying of cancer in Ontario, Canada. PATIENTS AND METHODS: Hospital separation records were linked to a population-based cancer registry to identify factors associated with hospitalization in the 203,713 patients who died of cancer in Ontario between 1986 and 1998. RESULTS: Between 1986 and 1998, 5.3% of all acute care beds in Ontario were devoted to the care of cancer patients in the last 6 months of life. The mean time spent in hospital in the last 6 months of life decreased from 34.3 days in 1986 to 22.7 days in 1998. Hospitalization rates increased exponentially during the last month of life. Patients younger than 50 years of age, women, and residents of poorer communities spent significantly longer in hospital than others. Hospitalization rates differed very little among the common solid tumors, but patients with CNS malignancies, the lymphomas, and the leukemias spent significantly longer in hospital than the other groups. There was significant interregional variations in hospitalization that were not explained by differences in case mix. There was a statistically significant inverse correlation between the rate of use of palliative radiotherapy and the hospital bed use in the county in which the patient resided. CONCLUSION: The total time spent in hospital in the last 6 months of life has decreased over the last decade, but acute care hospitals continue to play a large role in the care of patients who are dying of cancer.
CANCER IS ONE OF THE leading causes of death in most developed countries. In the United States and Canada, it presently accounts for approximately one quarter of all deaths, and it is expected that the number of cancer deaths will continue to increase as the population grows older over the next decade.1,2 The increasing number of deaths from cancer has created increasing demands for supportive care and palliative treatment, and this places a significant added burden on our health care systems. Inpatient care in hospital is commonly believed to be the most expensive way of providing palliative care,3 but despite the growth of hospice and community-based home care, in some communities a large proportion of cancer deaths still takes place in acute care hospitals.4-7 Little is known, however, about hospital bed utilization in the months that precede death from cancer, when palliation is usually the main goal of care. Since the mid-1980s, health care restructuring across Canada has led to hospital closures and a reduction in the number of acute care beds.8,9 Health care providers have been pressured to shorten hospital stay and reduce admissions. This has resulted in substantial reductions in length of stay in Canadian neonatal and maternity units, for example, but its impact on cancer care is unknown.10,11 The purpose of this study was to describe hospital bed utilization in the last 6 months of life among patients who died of cancer in Ontario between 1986 and 1998 and to identify patient-related, disease-related, and health system–related factors associated with hospitalization in this patient group.
Sources of Data The Ontario Cancer Registry (OCR) is a population-based registry that covers the province’s entire population of 11.6 million people. The OCR is a passive registry that receives information from four major sources: hospital separation records, clinical records from provincial cancer centers in Ontario, death certificates, and pathology reports from laboratories across the province. The OCR uses a system of probabilistic record linkage to identify cases and create a composite file that contains the following information: ICD9 site code, ICD_O histology code, date of diagnosis, date of birth, place of residence at diagnosis, vital status, date of death, and cause of death. The linkage procedures and completeness of cancer registration are well documented elsewhere.12,13 For the purpose of this study, hospital separation records were linked back to the OCR to provide the following additional information about hospital admissions: type of hospital, admission date, discharge date, admitting diagnosis, and other diagnoses and procedures. Over the study period, hospital participation in collection of separation records was consistent and complete throughout Ontario.14 The Canadian Census provided an electronic file that contained information about community socioeconomic status (SES), including median household income at the level of the census enumeration area and at the level of the census subdivision.15 As described previously, these data were linked to individual cases in the OCR, based on their place of residence, to provide an ecologic measure of socioeconomic status.16 Information about the number of hospital beds available and bed occupancy rates in Ontario was provided by the Ontario Hospital Association and Statistics Canada.9,17
Study Population
Outcome Variables
Study Variables
Analysis Multivariate analysis was done using the stepwise general linear regression model. The dependant variable in the model, HD6m, was logarithmically transformed to achieve a near-normal distribution. Percentage difference in hospital days between each subgroup and its reference group was calculated by the formula of (eB - 1) x 100%. Absolute difference was computed by least-square means method (SAS Institute Inc, Cary, NC).
Hospitalization Rate We identified 203,713 patients with a pathologically confirmed diagnosis of cancer who died of cancer in Ontario between 1986 and 1998. Figure 1 shows a frequency distribution of the total number of days spent in hospital in the last 6 months of life for the study population as a whole. The mean of the distribution was 34.1 days, and the median was 24 days. Only 7.7% of patients were never hospitalized in the last 6 months of life, 5.5% were hospitalized for 30 days or less, 24.7% for 31 to 60 days, 9.1% for 61 to 90 days, and 7.1% for more than 90 days.
Figure 2 shows time trends in the rate of hospitalization over the last six months of life. Figure 2A shows that the proportion of all cases hospitalized on any given day increases slowly until approximately 3 months before death, but then increases at an ever-increasing rate as death approaches. Figures 2B and 2C simultaneously illustrates time trends in hospitalization rates for selected subgroups of patients defined by primary site. The pattern observed in each of the major disease groups illustrated is fairly similar. Exceptions include patients with CNS malignancies and leukemia who have higher rates of hospitalization than other cases for several months before death. The hospitalization rates for these and all other diseases converge in the last month of life. Figure 2D illustrates changes in time trends in hospitalization rates over the 13-year period of this study. The proportion of cases that were in hospital on any given day during the last 6 months of life decreased steadily over the period of the study. The absolute difference in the proportion of patients hospitalized was greatest close to the end of life; the proportion of patients in hospital on the day of death decreased from 82.4% in 1986 to 59.0% in 1997 to 1998. However, the relative decrease in the proportion of patients hospitalized was greater several months before death. For example, 180 days before death, 13.4% of patients were in hospital in 1986, and this fell by more than half to 6.0% in 1998.
Over the entire study period, 94.8% of all admissions were to acute care hospitals, and the rest of admissions were to chronic care or rehabilitation facilities. Cancer patients in the last 6 months accounted for 5.3% of all acute care bed days used in Ontario. Figure 3 shows that the time spent in hospital per case decreased more or less in parallel with the decline in the total number of hospital beds available in Ontario. As a consequence, there was a net decrease in the total number of bed days used by cancer patients despite the increase in cancer mortality.
Factors Affecting Time Spent in Hospital Table 1 lists the mean and median number of days spent in hospital in the last 6 months of life (HD6m) for subgroups of patients defined by patient-related and disease-related characteristics. The youngest group of patients (< 50 years) spent longer time in hospital than any other age group, and that difference was statistically significant. Although the median time spent in hospital declined steadily with increasing age, there was no significant difference in the mean number of days spent in hospital among older groups above 50 years of age. The disparity between the medians and the means is explained by differences in the shape of the distributions of time spent in hospital among the different age groups. The proportion of patients who were never hospitalized was highest in the oldest group, and this was responsible for lowering the mean, but the oldest group also contained the highest proportion of cases that spent a long period in hospital (> 90 days), and this increased the mean (data not shown). Female patients spent longer periods in hospital than male patients. Patients from poorer communities spent longer periods in hospital than those from wealthier communities. Although the sex and SES associations were statistically significant, they are relatively small.
Disease-related factors also influenced total time spent in hospital in the last 6 months of life. HD6m ranged from 51.4 days for primary CNS malignancies to 29.8 days for lung cancer. As might be expected, patients with a record of CNS or bone metastasis spent considerably longer in hospital than others. In contrast, a record of metastasis to the liver was not associated with a higher number of days spent in hospital. Records of the presence or absence of metastasis were not available in the 8.7% of patients who died without hospitalization between 1991 and 1998, which biases the main analysis toward finding an association between metastasis and the numbers of hospitalized days. However, a secondary analysis in which these cases were excluded produced similar results (data not shown). Predictably, patients with malignant spinal cord compression or pathologic fracture of the femur spent longer periods in hospital than other patients. Figure 4 illustrates geographic variation in HD6m across the 46 counties of Ontario for which reliable data were available. The greatest use of hospitalization in the last 6 months of life is observed in northern counties that are relatively remote from the provincial cancer centers, but there were also marked variations in HD6m across southern Ontario. Figure 5 illustrates the relationship between time spent in hospital in each county and the rate of palliative radiotherapy used here as an indicator of access to specialized care in the ambulatory care sector. There is a statistically significant inverse correlation between the rate of use of palliative RT measured at county level and the mean number of days spent in hospital during the last 6 months of life, also measured at county level (Pearson correlation coefficient r = .5, P < .01).
Table 2 shows that there was a steady decline in the total number of days that patients spent in hospital during the last 6 months of life over the 13-year period of this study; the mean of HD6m decreased by 34% from 34.3 days in 1986 to 22.7 days in 1998. Table 2 also shows that the reduction in total time spent in hospital was primarily attributable to a decrease in mean length of stay per admission, which decreased by 28.6% from 18.9 to 13.5 days, whereas the number of admissions decreased by only 7.7% from 1.82 to 1.68 admissions per case.
Multivariate Analysis All study variables except for metastatic disease, on which information was not available from before January 1991, were included in a general linear regression model, the results of which are summarized in Table 3. There were small but statistically significant differences in time spent in hospital as a function of age, sex, and SES. The primary site was associated with variation of as much as 14 days in the mean time spent in hospital, with all but breast being statistically significantly higher than lung. There was a statistically significant inverse association between hospitalization and the rate of the use of palliative RT in the county in which the patient resided. However, there were differences of as much as 6 days among the referral areas of the provincial cancer centers that were not explained by variations in palliative RT. Also consistent with our findings in univariate analysis, the mean number of hospital days per case decreased by 4.8% annually over the study period; patients who died in 1998 spent an average 13.9 fewer days in hospital than those who died in 1986.
Readmission Rates We also examined the rate of readmission to hospital to determine whether shorter lengths of stay were associated with higher rates of readmission to hospital within 7 days of discharge. We found a significant inverse correlation between the average duration of admission at county level and the readmission rate (Pearson correlation coefficient r = .5, P < .01). However, although the mean duration of stay decreased over the period of the study, 7-day readmission rates actually fell from 14.1% (95% CI, 13.7% to 14.5%) in 1986 to 10.4% (95% CI, 10.1% to 10.6%) in 1998.
To our knowledge, this is the first population-based study to describe the use of hospital care among patients with advanced cancer. We found that hospital care continues to play a large role in meeting the needs of patients who are dying of cancer. Only a minority of patients died without requiring some hospital care, and the average patient spent approximately 5 weeks in hospital during the last 6 months of life. With few exceptions, the temporal pattern of hospital bed utilization during the last months of life was remarkably similar among patients with different primary malignancies; the terminal phase of the disease is always associated with an exponential rise in hospitalization rates, presumably reflecting a steep increase in the need for medical and supportive care as death approaches. From the perspective of the health system, patients who are dying of cancer make important demands on the total resources of the hospital sector, accounting for approximately 5% of all hospital bed use in Ontario. Some characteristics of the patient and disease were predictably associated with the length of time that individual patients spent in hospital. Patients with disease states that cause prolonged disability, such as primary or secondary cancers involving the CNS, spent more time than average in hospital in their last 6 months of life. Patients with diseases that are usually managed with intensive systemic therapy, such as the leukemias and the lymphomas, also spent more time in hospital. We found that women spent rather more time in hospital than men, probably because women are more likely to be widowed and live alone and perhaps also because women are more willing and able to take on the role of providing care at home for their partners.19,20 Residents of poorer communities spent more time in hospital than residents of richer communities. This is in contrast to our previous finding that residents of poorer communities who are dying of cancer in Ontario are less likely to receive palliative RT than residents of richer communities.18 Taken together, these findings are consistent with other studies that show that poor patients generally receive less specialized care but more general medical care than richer members of the same community in countries with universal health insurance schemes.21-24 In addition to these fairly predictable associations, we found a significant temporal trend toward lower hospitalization rates and marked interregional variations in bed utilization, neither of which were explained by differences in case mix. The downward trend in the use of hospital care for patients who are dying of cancer was probably a consequence of the decrease in the total number of hospital beds available in Ontario, but the reasons for the geographic variations in bed use are less clear. It seems logical that hospital bed use would be influenced by the availability of beds or by the availability of the resources needed to provide comparable care in the community setting.25 Unfortunately, we did not have access to information about the availability of hospital beds or community care services at the regional level, so we could not test that hypothesis. It is known, however, that rural Ontario has difficulty recruiting and retaining the family physicians who are essential in providing adequate outpatient care for terminal cancer patients, and this probably contributed to the higher rates of hospitalization in the observed remote areas. We did find some evidence that access to specialized cancer care in the ambulatory care sector was associated with lower rates of hospital bed utilization. There was a significant inverse correlation between rates of radiotherapy utilization in the county in which the patient resided and total time spent in hospital. This is consistent with the hypothesis that active treatment for the debilitating symptoms of advanced cancer in the ambulatory care setting may reduce the need for hospitalization, although it may not be access to the radiotherapy itself that lowered hospital utilization. The regional cancer centers that provide radiotherapy in Ontario also provide a fairly comprehensive range of other treatments and supportive care services that may also be involved in lowering hospitalization rates. The greatest limitation of this study is that we had no direct information about the patients’ functional status, quality of life, or satisfaction with their care. We were, therefore, unable to relate differences in hospital use to health outcomes, and we cannot reach any firm conclusions as to whether the geographic variations in hospitalization rates were appropriate or the temporal decline in hospitalization was desirable. We did, however, measure readmission rates, which have been used as an indicator of quality of care in other settings. As might be expected, when we compared different regions, we found that shorter lengths of stay were significantly associated with higher early admission rates. We did not, however, find evidence of any increase in readmission rates as length of stay declined over the 13 years of the study. On the contrary, readmission rates decreased significantly. This does not imply that the shorter hospital admissions were achieved without any adverse impact on the patient, but it does suggest that the crisis responsible for the original admission was being dealt with equally effectively during the shorter hospital stays observed in the latter phase of the study. This is consistent with the reports from other Canadian provinces that hospital cutbacks in general have not been associated with higher readmission rates or indeed with poorer self-reported levels of health.8 There is certainly evidence that spending less time in hospital is desirable from the patient’s point of view,26-28 but that presupposes that the care they need can be provided equally well in the community. A decrease in hospital bed utilization also seems attractive from an economic perspective, but cutting hospital costs will not necessarily translate into overall savings in health care expenditure. The cost of the community-based care may prove to be equally high once the total cost of home care and outpatient services has been taken into account.27 In the present study, we were unable to determine whether lower rates of hospital utilization were associated with increased use of community supportive care services by cancer patients, although over the study period, there have been reports of increases in the number of visits to hospital emergency rooms, clinics, and other outpatient programs in the population at large.8,9 The major strength of this study is that it describes the use of hospital beds in the care of patients dying of cancer in the population at large. Previous studies that have addressed this issue have been based on patients referred to palliative or terminal care support programs, and their results are not necessarily applicable to the population as a whole.29 Our approach also enabled us to measure the hospitalization rates over the last 6 months of life, which is more informative than reporting only whether the patient died in hospital or not, the only information usually provided in reports about the final phase of life from palliative care services.7,30 Additional research is needed to assess the impact of alternative models of care on the quality of life of patients with advanced cancer and on the total cost of care. Ontario is now collecting standardized data about home care services, which may in the future permit a more complete analysis of the care of the dying in our community.8
Supported in part by grants from the National Cancer Institute of Canada and Cancer Care Ontario (to W.J.M.). The authors thank Eric Holowaty, MD, and the Ontario Cancer Registry for providing access to their data and for much helpful advice and Beverley J. Shortt for her skill and patience in preparation of the manuscript.
1. National Cancer Institute of Canada: Canadian Cancer Statistics 1998. Toronto, Canada, The Institute, 1998 2. Devesa SS, Grauman DJ, Blot WJ, et al: Atlas of cancer mortality in the United States: 1950-94. Washington, DC, United States Government Printing Office, 1999, Publication No. 99-4564 3. Mor V, Kidder D: Cost savings in hospice: Final results of the National Hospice Study. Health Serv Res 20: 407-422, 1985[Medline] 4. National Center for Health Statistics: Vital Statistics of the United States, 1989, II: Mortality. Hyattsville, MD, National Center for Health Statistics, 1993 5. MacDonald N: Cancer centers: Their role in palliative care. J Palliat Care 8: 38-42, 1992 6. Clifford CA, Jolley DJ, Giles GG: Where people die in Victoria. Med J Aust 155: 446-456, 1991[Medline]
7.
Bruera E, Hanson J, Selmser P, et al: Edmonton regional palliative program: Impact on patterns of terminal cancer care. Can Med Assoc J 161: 290-293, 1999 8. Canadian Institute for Health Information: Health Care in Canada: A First Annual Report. Http://www.cihi.ca 9. Statistic Canada: Health Statistics at a Glance 1999: Health Resources. Utilization, Status, Determinants, Causes of Death. Ottawa, Canada, Industry Canada, 1999, Publication No. 82F0075XCB 10. Wen SW, Liu S, Fowler D: Trends and variation in neonatal length of in-hospital stay in Canada. Can J Pub Health 89: 115-119, 1998[Medline] 11. Wen SW, Liu S, Marcoux S, et al: Trends and variations in length of hospital stay for childbirth in Canada. Can Med Assoc J 158: 875-880, 1998[Abstract] 12. Clarke EA, Marrett LD, Krieger N: Cancer registration in Ontario: A computer approach, in Jenson O, Parkin DM, MacLennan R, et al (eds): Cancer Registration Principles and Methods. Lyons, France, International Agency for Research on Cancer, 1991, pp 246-257 13. Robles SC, Marrett LD, Clarke EA, et al: An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 41: 495-501, 1988[CrossRef][Medline] 14. Williams JI, Young W: A summary of studies on the quality of health care administrative databases in Canada, in Goel V, Willims JI, Anderson GM, et al (eds): Patterns of Health Care in Ontario: The ICES Practice Atlas (ed 2). Ottawa, Canada, Canadian Medical Association, 1996, pp 339-345 15. Statistic Canada: Census Handbook 1996. Ottawa, Canada, Industry Canada, 1997, Publication No. 92-352-XPE 16. Mackillop WJ, Zhang-Salomons J, Groome PA, et al: Socioeconomic status and cancer survival in Ontario. J Clin Oncol 15: 1680-1689, 1997[Abstract] 17. Ontario Hospital Association: Greater accountability and financial management in the new health economy. Http://www.oba.ca
18.
Huang J, Zhou S, Groome P, et al: Factors affecting the use of palliative radiotherapy in Ontario. J Clin Oncol 19: 137-144, 2001 19. Perry GR, Roades de Meneses : Cancer patients at home: Needs and coping styles of primary caregivers. Home Healthcare Nurs 6: 27-30, 1989
20.
Katz SJ, Kabeto M, Langa KM: Gender disparities in the receipt of home care for elderly people with disability in the United States. JAMA 284: 3022-3027, 2000 21. Verrilli DK, Berenson R, Katz SJ: A comparison of cardiovascular procedure use between the United States and Canada. Health Serv Res 33: 467-487, 1998[Medline]
22.
Alter DA, Naylor CD, Austin P, et al: Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med 341: 1359-1367, 1999 23. Roos NP, Mustard CA: Variation in health and health care use by socioeconomic status in Winnipeg, Canada: Does the system work well? Yes and no. Milbank Q 75: 89-111, 1997[CrossRef][Medline] 24. Gutzwiller F, La Vecchia C, Levi F, et al: Education, disease prevalence and health service utilization in the Swiss National Health Survey "SOMIPOPS." Prev Med 18: 452-459, 1989[CrossRef][Medline]
25.
Tolle SW, Rosenfeld AG, Tilden VP, et al: Oregon’s low in-hospital death rates: What determines where people die and satisfaction with decisions on place of death? Ann Intern Med 130: 681-685, 1999 26. McWhinney IR, Bass MJ, Orr V: Factors associated with location of death (home or hospital) of patients referred to a palliative care team. Can Med Assoc J 152: 361-367, 1995[Abstract] 27. Gray D, MacAdam D, Boldy D: A comparative cost analysis of terminal cancer care in home hospice patients and controls. J Chronic Dis 40: 801-810, 1987[CrossRef][Medline] 28. Costantini M, Camoirano E, Madeddu L, et al: Palliative home care and place of death among cancer patients: A population-based study. Palliat Med 7: 323-331, 1993[Medline]
29.
Christakis NA, Escarece JJ: Survival of medicare patients after enrollment hospice programs. N Engl J Med 335: 172-178, 1996 30. Hunt R, Bonett A, Roder D: Trends in the terminal care of cancer patients: South Australia, 1981-1990. Aust N Z J Med 23: 245-251, 1993[Medline] Submitted October 11, 2000; accepted November 19, 2001.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
), acute care beds available (
), total number of bed days used by cancer patients in the last 6 months of life (
), and mean number of bed days per case (
) are shown as a percentage of 1986 values.
