Trends in the Aggressiveness of End-of-Life Cancer Care in the State of Qatar

Purpose Quality of end-of-life (EOL) care is a key component of excellence in cancer care, and monitoring indicators for quality of EOL cancer care is crucial to providing excellent care. The aim of the current study is to describe the relative aggressiveness of EOL cancer care in the state of Qatar and to compare it with international figures. Methods We analyzed all deaths from cancer in Qatar between January 1, 2009 and December 31, 2013. A total of 784 eligible patients were studied to assess aggressiveness of cancer care at EOL. Results The average number of intensive care unit admissions per person decreased from 0.44 to 0.22 (P < .001) over the period of study. In addition, patients spent fewer days in the intensive care unit (2.79 to 1.82 days; P = .006) and made fewer visits to the emergency department (1.00 to 0.52 visits; P < .001) in the last 30 days of life. Fewer patients had at least one aggressive treatment measure at EOL during the 5-year period (82.3% to 71.0%; P = .038). The mean composite score for aggressiveness of EOL care decreased from 2.24 to 1.92 (P < .01). Conclusion The aggressiveness of EOL cancer care has significantly decreased over time in Qatar; however, despite this decrease, the rate is still higher than that reported internationally. The integration of community palliative care services in Qatar may further decrease the aggressiveness of cancer care at EOL.


INTRODUCTION
Despite advances in early detection, treatment, and survival, cancer remains a leading cause of death in most developed countries. With the number of deaths from cancer and other chronic diseases expected to increase as a result of a growing and aging population, the need for quality end-of-life (EOL) care has become increasingly important. 1 During the past decade, cancer care at EOL has become increasingly aggressive, 2,3 resulting in disproportionately greater health expenditures for decedents than for survivors. [2][3][4][5] This increased treatment intensity, however, does not correspond to improved health outcomes and patient or family satisfaction, 6,7 and raises concerns about the appropriateness of providing aggressive care before patients die. Furthermore, the emotional burden, loss of hope after experiencing the failure of life-sustaining treatments, and financial costs can exact a substantial toll from patients, families, and society. 8,9 To systematically examine the quality of cancer care at EOL, Earle et al 10 developed major indicators of poor-quality care. These indicators are a combination of measures of overly aggressive care and underuse of supportive care services, such as palliative care or hospice services. Indicators are defined as the last dose of chemotherapy received within 14 days of death, a new chemotherapy regimen starting less than 30 days before death, more than one emergency department (ED) visit within 30 days of death, more than one hospital admission or spending more than 14 days in hospital in the last month of life, an intensive care unit (ICU) admission in the last month of life, or death in an acute care hospital. These indicators were adopted by researchers from several countries worldwide to illustrate the quality of care received in the last months of life by patients with cancer. [11][12][13][14][15] ; however, such patterns of cancer care have never been explored in the Middle East. The significance of evaluating such trends in the state of Qatar is highlighted by three facts: cancer accounted for 18% of all deaths in Qatar from 2000 to 2012 15 ; improving the quality of EOL cancer care is now recognized as a key factor for excellence in cancer care as a result of the Qatar universal health care system; and trends in the quality of cancer care at EOL can be analyzed in a population that encompasses all age and cancer groups, which facilitates comparison with international trends. These findings can be directly applied to inform national public health policies.
The purpose of this study, therefore, was to evaluate the aggressiveness of EOL cancer care in Qatar. A secondary end point was to compare our results with those observed worldwide.

Data Sources
All deaths between 2009 and 2013 for which cancer was the underlying cause were extracted from the National Registry of Deaths Database. By law in Qatar, a death must be registered and the underlying cause of death recorded to approve the deceased for burial. The underlying cause of death was recorded in the database using the International Classification of Diseases, Tenth Revision, codes.

Study Design and Study Population-Cohort Selection
This retrospective, population-based cohort study analyzed all adult cancer deaths in Qatar registered by the National Registry of Deaths Database from January 1, 2009 to December 31, 2013.
A retrospective chart review was conducted to measure for markers of aggressive EOL cancer care, defined as the occurrence of any of the following indicators: the last dose of chemotherapy received within 14 days of death, a new chemotherapy regimen started less than 30 days before death, more than one ED visit within 30 days of death, more than one hospital admission or spending more than 14 days in the hospital in the last month of life, an ICU admission in the last month of life, or death in an acute care hospital.

Explanatory Variables
Explanatory variables included the following: age at death (reported as a continuous variable), gender (male v female), nationality (Qatari v other nationality), primary tumor site (lung, breast, colorectal, lymphoma, brain, prostate, or other), and place of death (palliative care, oncology unit, Hamad General Hospital; medical ICU; accident/emergency; brought-in-dead unit; and other).

Main Outcome Measure
We developed indicators of aggressive care by identifying cutoffs rounded to clinically meaningful numbers. The following indicators were measured as described elsewhere 10 : started a new chemotherapy regimen in the last month of life, received the last dose of chemotherapy within 14 days of death, had more than one ED visit within 30 days of death, spent more than 14 days in the hospital at last encounter, had more than one hospitalization within 30 days of death, or had at least one ICU admission within 30 days of death.
We then created the main outcome measure, a categorical composite measure of aggressive EOL care, which was defined as the occurrence of at least one of the indicators of aggressive care. The categorical composite measure of aggressive EOL care was reported as the proportion of aggressiveness of care.

Statistical Analyses
Descriptive statistics were summed as means and standard deviation of age and calculated as percentages for gender, nationality, primary site of cancer, and place of death. We then conducted an analysis of variance to test for differences in the mean age by year as well as a x 2 test to test for differences in proportions of the categorical variables by year. The Cochrane-Armitage test was used to determine significance of trends over time.
The Cochran-Armitage test was used to evaluate time trends, from 2009 to 2013, of the categorical composite measure of aggressiveness. A continuous composite measure of aggressive EOL care was reported as the average number of aggressive care measures by creating a score of those measures with a minimum score of not having any measure of aggressiveness and a maximum score of having all five measures of aggressiveness. Spearman rank-order correlation was used to evaluate time trends of the continuous composite measure of aggressive EOL. Results are reported in Table 1.
Multivariable logistic regression was used to examine the simultaneous effects of year of death, age, gender, nationality, primary tumor site, and place of death on the categorical composite measure of aggressive EOL care, that is, patients who experienced at least one indicator of aggressive care. Unadjusted odds ratios (UORs) and adjusted odds ratios (ORs) and corresponding 95% CIs were computed and are reported in Table 2.
All statistical analyses were conducted using STATA software version 13 (STATA, College Station, TX; Computing Resource Center, Santa Monica, CA). P , .05 was considered significant.

Ethics and Permission
This study was approved by the Research Ethics Committee, Hamad Medical Corporation, Medical Research Center, Qatar. Table 3 shows the characteristics of expired patients with cancer during the 5-year study period of 2009 to 2013 who met the eligibility criteria, that is, patients who spent < 31 days in the hospital upon last admission (N = 784). Two hundred forty-one patients (23.5%) spent more than 31 days in the    cancer. There was no statistical significance by the year of study (P = .057). More patients died in the palliative care unit from 2011 on compared with earlier years, and a decreasing trend of death in the medical ICU by year of death was observed (P = .001).

Trends in Measures of Care and Indicators of Aggressive Care Among Patients With Cancer During the 5-Year Study Period
Of the 784 patients with cancer included in the study period, the average number of ICU admissions between 2009 and 2013 decreased from 0.44 to 0.22 (P , .001; Fig 1 and Table 4).  Table 4). From 2009 to 2013, fewer patients had more than one ED admission in the last month of life (22.2% v 7.7%, respectively; P , .001; Table 5) or were admitted to the ICU (42.9% v 21.9%, respectively; P , .001; Table 5). However, between 2009 and 2013, the average number of days spent in the hospital at last encounter increased from 7.72 to 9.64 days (P = .029; Fig 1 and Table 4).

Composite Measure of Aggressiveness by Year
Patients with at least one measure of aggressive care were considered to have been exposed to aggressiveness of care. The proportion of aggressiveness  of care significantly decreased during the 5-year study period from 82.3% of patients exposed to at least one measure of aggressiveness to 71.0% (P = .038; Table 1). In addition, the average number of aggressive care measures, varying from zero measures per patient to five measures per patient, significantly decreased from 1.33 measures in 2009 to 1.02 in 2013 (P = .010; Table 1). Table 2 shows UORs and ORs and corresponding 95% CIs for patients with cancer who experienced at least one measure of aggressive care in the last month before death. At the bivariable level, year of death was found to be a significant protective factor against aggressive care (UOR, 0.88; 95% CI, 0.78 to 0.99) as well as age of the patient (UOR, 0.99; 95% CI, 0.98 to 0.99) and being of Qatari nationality (UOR, 0.59; 95% CI, 0.42 to 0.82). No associations were found between gender or the primary site of cancer and aggressiveness of care. However, patients who died in the ED and the brought-in-dead unit were less likely to experience aggressive care (UOR, 0.21; 95% CI, 0.11 to 0.41; and UOR, 0.09; 95% CI, 0.05 to 0.16, respectively) compared with patients who died in the palliative care unit.

Multivariable Logistic Regression
Multivariable logistic regression shows that older patients compared with younger patients (OR, 0.98; 95% CI, 0.97 to 0.99) and patients who died in the ED and the brought-in-dead units compared with patients who died in the palliative care unit (OR, 0.18; 95% CI, 0.09 to 0.36; and OR, 0.07; 95% CI, 0.03 to 0.14, respectively) were less likely to experience aggressive interventions.

DISCUSSION
The current study showed a significant declining trend in the proportion of patients experiencing aggressive care at EOL, from 82% in 2009 to 71% in 2013 (P = .038). Despite this decline in the incidence of aggressiveness of care, 71% remains higher than percentages reported in the United States, Canada, Europe, and Asia. 5,7,9,16 This is most likely attributable to many factors, including a lack of chronic care facilities (eg, hospice or home health care) to which patients can be referred when no acute symptoms require hospitalization as well as the lack of appropriate prognostication models that can accurately predict whether a patient will die within days, weeks, or months. In addition, delayed referral to palliative care and the inherent unease among oncology and hematology physicians over referring patients to palliative care deprives patients of a quality of life that focuses on important concerns at an advanced stage of disease, such as pain control, psychosocial support, and relief of spiritual distress. Aggressiveness of care at EOL has its known negative effect on patient quality of life as well as its detrimental effect on patient families subjected to a significantly higher rate of psychologic distress compared with those whose relatives or loved ones are not subjected to aggressive management during the last month of life. 1,6 The decline in aggressive care at EOL in the current study is mostly a result of the opening of the palliative care unit at the end of 2008, the implementation of a do not attempt resuscitation (DNAR) policy in 2011, and the addition of a modified National Comprehensive Cancer Network screening tool for hospitalized patients.
The DNAR policy of our hospital states that DNAR is a clinical decision made by the primary physician and two other oncology physicians. It is not presented to patients or families as a choice; rather, they are informed of the decision. This of course has limited unnecessary intubation and ICU admission for patients with advanced cancer at EOL. The modified NCCCR palliative care referral scoring tool consists of six items, and patients who score a 5 must at least have a palliative care consultation, whereas patients scoring > 7 should be transferred to palliative care. These measures have led to an earlier referral of patients to palliative care beginning in 2012.
In conclusion, the current study shows a decline in the percentage of aggressive care administered to adult patients with cancer in the last month of life from 2009 to 2013. Despite this observed decline, the incidence of aggressive care at EOL in our institution remains significantly higher than that reported in Europe or North America.
Early referral to palliative care, more accurate prognostication near EOL, and availability of chronic care facilities to which patients can be referred should further decrease the percentage of patients who receive aggressive care at our facility.