Article Tools
Special Series: Palliative Care
ORIGINAL CONTRIBUTIONS
Article Tools
Impact of Palliative Medicine Involvement on End-of-Life Services for Patients With Cancer With In-Hospital Deaths
Abstract
Palliative care (PC) has been shown to improve the quality of care and resource utilization for inpatients. We examined the relationship between PC consultation before and during final admission and patterns of care for dying patients at our tertiary cancer center.
We retrospectively reviewed adult patients with solid tumor cancer with a length of stay ≥ 3 days who died in hospital between December 2012 and November 2014. We recorded services, including laboratory testing, imaging, blood products, medications, diet orders, do not resuscitate orders, and consultations, delivered within 3 days of death. We assessed the differences among services delivered to patients with outpatient PC, inpatient PC only, and no PC involvement.
Of 695 patients, 21% received outpatient PC, 46% received inpatient PC only, and 33% received no PC. During their final admission, 11.2% of patients received radiation therapy, and 12.5% received tumor-directed therapy, with no differences on the basis PC involvement (P = .09 to .17). In the last 3 days of life, imaging tests occurred in 50.1%; patients with outpatient or inpatient-only PC underwent fewer studies (43.5% and 47.3%) than did those with no PC involvement (58.1%; P = .048). Do not resuscitate orders were in place within the 6 months before final admission at a greater rate for patients with outpatient PC (22%) than for patients with inpatient-only PC (8%) or those with no PC involvement (12%; P = .002).
In this retrospective cohort of patients with solid tumor dying in hospital, few patients received cancer-directed therapies at the end of life. Involvement of PC was associated with a decrease in diagnostic testing and other services not clearly promoting comfort as patients approached death.
Palliative care (PC) involvement with the patient with cancer has been shown to improve the quality of life before death.1-3 Evidence of the influence of PC involvement on the place of death or resource use at the end of life (EOL) is mixed. Between one fifth and one half of patients with cancer continue to die in hospital,4-6 and some interventions including diagnostic testing and other services may not promote comfort at EOL.7 Previous studies have described the services delivered to patients with cancer during the last month8 or several months of life9 or focus on specific services such as chemotherapy near death,10,11 but none describe the use of more routine tests and treatments performed in hospitalized patients who are dying as a result of cancer. PC consultation during hospital admission has been shown to reduce EOL resource use in a general population, but the specific impact of outpatient PC consultation on the services delivered to dying inpatients with cancer is not clear.
We evaluated the services delivered to patients with solid tumors who died in our tertiary cancer center over a 2-year period. We focused on this population because their impending deaths were more likely to be anticipated by the medical staff. We examined the differences between services rendered to patients with outpatient PC consultation within the 6 months before their final admission, patients with PC consultation during their final admission only, and patients with no PC involvement. We hypothesized that involvement of PC before the final admission would be associated with lower use of these services compared with later (during final admission) PC involvement and no PC involvement.
With institutional review board approval, we used our institutional database to identify adult patients (18 years and older) with a solid tumor who died during admission between December 1, 2012, and November 30, 2014, at our tertiary cancer center. We limited our analysis to patients with a length of stay of ≥ 3 days to eliminate patients with rapid decline, in whom death may have been unexpected. We queried cancer registry, phlebotomy, order entry, pharmacy, and radiology databases to collect patient-level demographics and the timing of delivered services, including laboratory testing, imaging, blood products, medications, consultation services, total parenteral nutrition and tube feeding orders, do not resuscitate (DNR) orders, and diet nil per os (NPO) orders, occurring between hospital admission and death. Data on chemotherapy within 14 days of death regardless of admission status were also collected. Children, patients who died in intensive care or a surgical unit, and patients with primary hematologic malignancies were excluded to create a more homogeneous sample.
Medications were categorized on the basis of the foundational database of Multum Medisource Lexicon. We excluded those classified as general anesthetics, miscellaneous agents, and radiologic agents, and medications missing a categorization. Opioids and other analgesic medications were classified as CNS agents. Drugs appearing multiple times with different doses were collapsed into a single use of the drug. A detailed breakdown of medication categorizations is given in Appendix Table A1 (online only). All interventional radiology procedures were excluded from radiologic studies because these were considered procedures and not pure radiographic scans. Multiple blood draws from a single patient taken within 15 minutes were considered to be part of the same draw and were collapsed into one record.
We grouped consultations into MD, Physical or Occupational Therapy, Swallowing, Integrative Medicine, Nutrition, and Chaplaincy or Social Work. For PC consultations, patients were classified into the following three categories: inpatient PC only, if they had no outpatient PC exposure but were consulted during their final admission; none, if they never met with a PC practitioner; and outpatient PC, if they had had at least one outpatient PC consultation within the 6 months before the final admission. Patients in the outpatient PC category may or may not have also received an inpatient PC consultation.
We calculated the proportion of patients who had each service within 3 days before death, broken down by PC status. All patients had at least 3 days of admission, so the denominator was equal to the total sample size. We also calculated the proportion of patients who had chemotherapy within 14 days of death, regardless of admission status. We assessed differences in treatment rates using Fisher’s exact test. All P values were adjusted for multiple comparisons using the false-discovery rate correction, and adjusted P values < .05 were considered statistically significant.
For the last 14 days of admission, we determined the cumulative proportion of patients who had a DNR order on each day. The patients with a DNR order in place before 14 days were counted in the initial denominator. All analyses were performed using SAS 9.4 (SAS Institute, Cary, NC).
The patient sample included 695 patients with a median age of 64 years (range, 19 to 93 years). The sample was divided evenly between male (n = 348 [50%]) and female (n = 347 [50%]) patients. The majority identified as white (n = 488 [70%]); 11.5% (n = 80) were black, and 7.6% (n = 53) were Asian. Patients were largely insured through Medicare (n = 321 [46.2%]) or through commercial or private insurance (n = 317 [45.6%]). Patients were admitted for a median length of stay of 9 days (range, 3 to 160 days; Table 1). Of the 695 patients, 147 (21%) received outpatient PC consultation within the 6 months before admission, 319 (46%) received inpatient PC consultation only, and 229 (33%) had no PC exposure. Among the patients receiving inpatient PC, consultations were ordered a median of 6 days before death (range, 0 to 97 days).
|
Patients who had outpatient PC were generally younger (median, 57 years; range, 23 to 87 years) compared with patients with inpatient PC only (median, 65 years; range, 20 to 93 years) and patients with no PC (median, 68 years; range, 19 to 90 years; P < .001). No differences in PC involvement were seen on the basis of ethnicity (P = .54) or sex (P = .26; Table 1).
Patients with inpatient PC only had the longest length of stay (median, 12 days; range, 3 to 160 days) compared with patients with outpatient PC (median, 8 days; range, 3 to 89 days) or those with no PC (median, 7 days; range, 3 to 74 days; P < .001).
Within 3 days before death, 541 patients (77.8%) had laboratories drawn; 334 (48.1%) had venipuncture (VP) and 413 (59.4%) had blood drawn without VP (eg, from an indwelling catheter). No differences on the basis of PC involvement were seen in the proportion of patients who received blood draws overall (P = .39) or non-VP blood draws (P > .95). A lower proportion of patients with outpatient PC (42.2%) had VP blood draws compared with those with inpatient PC only (45.1%) or those with no PC (55.9%; P = .050; Table 2).
|
Within 3 days before death, 114 patients (16.4%) in the full cohort received blood products. Twenty-four patients with outpatient PC (16.3%), 58 patients with inpatient PC only (18.2%), and 32 patients without PC (14.0%) received blood products (P = .55; Table 2).
Within 3 days before death in the full cohort, 358 patients (50.1%) received imaging tests. X-rays and computed tomography scans were the most common studies; a few patients underwent positron emission tomography scanning (Table 2). Overall, a lower proportion of patients with outpatient PC (43.5%) received imaging tests compared with those with inpatient PC only (47.3%) or no PC (58.1%; P = .048). This difference was driven largely by a difference in receipt of x-ray studies (P = .051), with 38.1%, 41.1%, and 52.0% of patients with outpatient PC, inpatient PC only, and no PC receiving x-rays, respectively.
Within 3 days before death, 79 patients (11.4%) had NPO orders in place. No differences were seen in the proportion of NPO patients who had received outpatient PC (9.5%), inpatient PC only (10.3%), or no PC consultation (14%; P = .49). Six patients (0.9%) received total parenteral nutrition, and 22 patients (3.2%) received tube feedings. No differences were noted on the basis of PC consultation (P = .28 to .85).
Within 3 days before death, 29 patients (4.2%) received radiation, and 30 patients (4.3%) received chemotherapy. There were no significant differences on the basis of PC involvement in the rates of radiation (P = .10) or chemotherapy (P = .19) within 3 days before death. Within 14 days before death, 113 patients (16.3%) received chemotherapy regardless of admission status. No differences were seen on the basis of PC involvement (Table 3).
|
All patients had medications administered within 3 days before death; the most common medications were CNS agents (694 of 695 [99.9%]), respiratory agents (675 of 695 [97.1%]), GI agents (675 of 695 [97.1%]), topical medications (517 of 695 [74.4%]), and anti-infectives (493 of 695 [70.9%]; Table 3). A lower proportion of patients with outpatient PC involvement (41.5%) received coagulation modifiers compared with patients with inpatient PC only (46.7%) and those without PC (59.8%; P = .005). Similarly, a lower proportion of patients with outpatient PC (4.8%) received antihyperlipidemic agents compared with patients with inpatient PC only (6.6%) and those without PC (13.5%; P = .024). More patients with outpatient (55.8%) and inpatient (56.4%) PC consultations received psychotherapeutic agents compared with patients without PC involvement (39.7%; P = .002). No other significant differences on the basis of PC involvement were seen for other medication categories (P = .08 to > .95; Table 3).
Within 3 days before death, 279 patients (40.1%) received consultations, of which the plurality was physician specialist consultations (47.3%). Patients with outpatient PC involvement had the lowest rate of MD consultations (11.6%) compared with those with inpatient PC only (19.7%) or no PC (22.7%); however, this difference was not significant (P = .08). A higher proportion of patients with inpatient PC only (16.9%) had chaplaincy or social work consultations compared with those with outpatient PC (10.9%) or no PC (5.2%; P = .001). No differences were seen among patients receiving outpatient, inpatient only, or no PC for physical or occupational therapy (10.2% v 11.6% v 18.3%, respectively; P = .10), swallowing (2.7% v 5.3% v 4.8%, respectively; P = .55), or integrative medicine (7.5% v 6.3% v 4.4%, respectively; P = .55) consultations.
A greater proportion of patients with outpatient PC had a DNR order in place within the 6 months before admission (22%) compared with patients with inpatient PC only (8%) or those with no PC involvement (12%; = .002). During the final admission, the cumulative proportion of patients with DNR orders rose exponentially, from 27.1% of patients on day 14 before death to 92.1% by the day of death. Figure 1 illustrates the trend in DNR orders on the basis of PC consultation status.
Fig 1. Cumulative proportion of patients with do not resuscitate (DNR) orders per day on the basis of palliative medicine consultation status. This figure represents the cumulative percentage of patients with a DNR order in place by a given day relative to their own death. Each percentage includes those from the previous days plus those who had a DNR order placed on that day. All percentages include those who had a DNR order in place before 14 days of death. This figure demonstrates that patients with outpatient palliative consultations started with a higher proportion of DNR orders, followed by patients with inpatient-only palliative consultation. In addition], this figure demonstrates the exponential trend of increased DNR orders as death approaches.
In our retrospective cohort of patients with solid tumors dying in the hospital, prior involvement of PC specialists in the outpatient setting was associated with lower use, during the last 3 days of the final hospitalization, of services that may not have promoted the patients’ comfort or other goals at this critical time. Such services included diagnostic testing by VP blood draw and x-ray studies, which often require transport of patients around the hospital, as well as administration of anticoagulants and statins.12 Patients who received outpatient PC or inpatient PC only were more likely to receive psychotherapeutic agents, including antidepressants, which may have improved the patient experience at EOL,13 and inpatient PC consultation strengthened the interdisciplinary support for patients through increased involvement of social work and chaplaincy, which also enhances quality of life.14-17 These associations were observed in a cohort of patients of a median younger age; EOL care is often more intensive in younger patients.18
Regardless of specialist PC involvement, nearly all patients in this cohort received analgesic medications at EOL, suggesting that primary and subspecialty PC addressed the key concern expressed by dying patients for relief of their pain.19,20 Chemotherapy administration was rare among all groups, reflecting recognition of the limited benefit of continued antineoplastic treatment of patients with solid tumors who are approaching death.10,21 Diagnostic testing was more common, with the majority of patients undergoing blood draws and imaging tests during the last 3 days of life. Whether the potential benefits of such testing outweighed the potential discomforts or other harms to patients cannot be determined from this analysis. We also cannot determine the net benefit or harm of coagulation modifiers, administered to almost one half of all patients in the last 3 days of life, often by subcutaneous injection, anti-infectives, nutritional agents (eg, vitamins), and blood products. In some situations, such services may have promoted comfort and been consistent with patient goals. In others, continuing re-evaluation of potential harms, including patient distress, and clarification of goals of care may support a decision to discontinue services not clearly promoting the patient’s comfort, especially when those services use scarce resources.23-25
There are several possible explanations for a greater impact of earlier, outpatient PC consultation, compared with later, in-hospital PC consultation, on the delivery of certain services. First, PC consultation often entails addressing patient and family coping with, understanding of, and education about the illness.26,27 These communication interventions may enhance patients’ and families’ prognostic awareness.28,29 Recognition of life-limiting illness earlier in the disease course, and in the outpatient setting, may help patients and their families share in proactive medical decision making that prioritizes comfort at EOL.30-32
In the group of patients receiving inpatient PC consultations only, these consultations were requested a median of 6 days before death for patients with a median hospitalization of 12 days. Late referral of patients with longer and likely complicated admissions may have limited the impact of inpatient PC.
Our study has several limitations. First, the study was retrospective and we cannot distinguish the net benefit or harm to the patient of specific services. We also cannot identify the point at which the patient was considered by his or her clinician to be actively dying. However, prognostication in patients with solid tumors has been well described.33,34 Second, given the nature of our study, we were unable to capture nuanced patient information such as the level of consciousness or cancer stage at the time of admission; some of these factors may have differed among PC groups and may have influenced care. In addition, we did not control for the degree of PC involvement in either of the PC groups, which ranged from one visit to many visits. It is possible that some patients were newer to our system than others and had less opportunity to receive outpatient PC during the 6 months leading to the index admission. These factors may have led us to underestimate the impact of close involvement with PC. Our study reflects care at a single cancer center and the findings may not be generalizable to other centers, locations, or populations. Finally, we do not know why these particular patients had in-hospital deaths or whether hospice was involved in their care before admission. Given that rates of in-hospital cancer deaths remain substantial,4-6 our findings are relevant to a large population of patients.
In this retrospective cohort of patients with solid tumors dying in hospital, most patients received medications addressing comfort needs and did not receive cancer-directed treatments as they approached death. Involvement of PC was associated with a decrease in diagnostic testing and other services that were not clearly promoting patient comfort. Future studies should prospectively address the impact of current practice, including the involvement of PC specialists in outpatient and inpatient settings, on patient quality of life and other outcomes as patients approach death.
Conception and design: All authors
Collection and assembly of data: Alison Wiesenthal, Debra A. Goldman
Data analysis and interpretation: All authors
Manuscript writing: All authors
Final approval of manuscript: All authors
Accountable for all aspects of the work: All authors
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/journal/jop/site/misc/ifc.xhtml.
No relationship to disclose
No relationship to disclose
Honoraria: Vedanta Biosciences (I)
ACKNOWLEDGMENT
Supported by Cancer Center Support Grant P30 CA008748 from the National Cancer Institute to Memorial Sloan Kettering Cancer Center. Presented in abstract form at the Lown Institute 4th Annual Conference, Chicago, IL, April 15, 2016, and at the 2015 Palliative Care in Oncology Symposium, Boston, MA, October 9-10, 2015.
| 1. | Higginson IJ, Evans CJ: What is the evidence that palliative care teams improve outcomes for cancer patients and their families? Cancer J 16:423-435, 2010 Crossref, Medline, Google Scholar |
| 2. | Rabow M, Kvale E, Barbour L, et al: Moving upstream: A review of the evidence of the impact of outpatient palliative care. J Palliat Med 16:1540-1549, 2013 Crossref, Medline, Google Scholar |
| 3. | Temel JS, Greer JA, Muzikansky A, et al: Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med 363:733-742, 2010 Crossref, Medline, Google Scholar |
| 4. | Barbera L, Seow H, Sutradhar R, et al: Quality of end-of-life cancer care in Canada: A retrospective four-province study using administrative health care data. Curr Oncol 22:341-355, 2015 Crossref, Medline, Google Scholar |
| 5. | Bekelman JE, Halpern SD, Blankart CR, et al: Comparison of site of death, health care utilization, and hospital expenditures for patients dying with cancer in 7 developed countries. JAMA 315:272-283, 2016 Crossref, Medline, Google Scholar |
| 6. | Hunt KJ, Shlomo N, Addington-Hall J: End-of-life care and achieving preferences for place of death in England: Results of a population-based survey using the VOICES-SF questionnaire. Palliat Med 28:412-421, 2014 Crossref, Medline, Google Scholar |
| 7. | Morrison RS, Ahronheim JC, Morrison GR, et al: Pain and discomfort associated with common hospital procedures and experiences. J Pain Symptom Manage 15:91-101, 1998 Crossref, Medline, Google Scholar |
| 8. | Wu E, Rogers A, Ji L, et al: Escalation of oncologic services at the end of life among patients with gynecologic cancer at an urban, public hospital. J Oncol Pract 11:e163-e169, 2015 Link, Google Scholar |
| 9. | Jang RW, Krzyzanowska MK, Zimmermann C, et al: Palliative care and the aggressiveness of end-of-life care in patients with advanced pancreatic cancer. J Natl Cancer Inst 107:dju424, 2015 Crossref, Medline, Google Scholar |
| 10. | Campion FX, Larson LR, Kadlubek PJ, et al: Advancing performance measurement in oncology: Quality oncology practice initiative participation and quality outcomes. J Oncol Pract 7:31s-35s, 2011 Link, Google Scholar |
| 11. | Lopez-Acevedo M, Havrilesky LJ, Broadwater G, et al: Timing of end-of-life care discussion with performance on end-of-life quality indicators in ovarian cancer. Gynecol Oncol 130:156-161, 2013 Crossref, Medline, Google Scholar |
| 12. | Kutner JS, Blatchford PJ, Taylor DH Jr, et al: Safety and benefit of discontinuing statin therapy in the setting of advanced, life-limiting illness: A randomized clinical trial. JAMA Intern Med 175:691-700, 2015 Crossref, Medline, Google Scholar |
| 13. | Rosenstein DL: Depression and end-of-life care for patients with cancer. Dialogues Clin Neurosci 13:101-108, 2011 Medline, Google Scholar |
| 14. | Hearn J, Higginson IJ: Do specialist palliative care teams improve outcomes for cancer patients? A systematic literature review. Palliat Med 12:317-332, 1998 Crossref, Medline, Google Scholar |
| 15. | McClain CS, Rosenfeld B, Breitbart W: Effect of spiritual well-being on end-of-life despair in terminally-ill cancer patients. Lancet 361:1603-1607, 2003 Crossref, Medline, Google Scholar |
| 16. | Miller JJ, Frost MH, Rummans TA, et al: Role of a medical social worker in improving quality of life for patients with advanced cancer with a structured multidisciplinary intervention. J Psychosoc Oncol 25:105-119, 2007 Crossref, Medline, Google Scholar |
| 17. | Phelps AC, Lauderdale KE, Alcorn S, et al: Addressing spirituality within the care of patients at the end of life: Perspectives of patients with advanced cancer, oncologists, and oncology nurses. J Clin Oncol 30:2538-2544, 2012 Link, Google Scholar |
| 18. | Ho TH, Barbera L, Saskin R, et al: Trends in the aggressiveness of end-of-life cancer care in the universal health care system of Ontario, Canada. J Clin Oncol 29:1587-1591, 2011 Link, Google Scholar |
| 19. | Singer PA, Martin DK, Kelner M: Quality end-of-life care: Patients’ perspectives. JAMA 281:163-168, 1999 Crossref, Medline, Google Scholar |
| 20. | Steinhauser KE, Christakis NA, Clipp EC, et al: Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 284:2476-2482, 2000 Crossref, Medline, Google Scholar |
| 21. | Levy MH, Back A, Benedetti C, et al: NCCN clinical practice guidelines in oncology: Palliative care. J Natl Compr Canc Netw 7:436-473, 2009 Crossref, Medline, Google Scholar |
| 22. | Reference deleted Google Scholar |
| 23. | Earle CC, Park ER, Lai B, et al: Identifying potential indicators of the quality of end-of-life cancer care from administrative data. J Clin Oncol 21:1133-1138, 2003 Link, Google Scholar |
| 24. | Ellershaw J, Ward C: Care of the dying patient: The last hours or days of life. BMJ 326:30-34, 2003 Crossref, Medline, Google Scholar |
| 25. | Reinbolt RE, Shenk AM, White PH, et al: Symptomatic treatment of infections in patients with advanced cancer receiving hospice care. J Pain Symptom Manage 30:175-182, 2005 Crossref, Medline, Google Scholar |
| 26. | Jacobsen J, Jackson V, Dahlin C, et al: Components of early outpatient palliative care consultation in patients with metastatic nonsmall cell lung cancer. J Palliat Med 14:459-464, 2011 Crossref, Medline, Google Scholar |
| 27. | Yoong J, Park ER, Greer JA, et al: Early palliative care in advanced lung cancer: A qualitative study. JAMA Intern Med 173:283-290, 2013 Crossref, Medline, Google Scholar |
| 28. | Jackson VA, Jacobsen J, Greer JA, et al: The cultivation of prognostic awareness through the provision of early palliative care in the ambulatory setting: A communication guide. J Palliat Med 16:894-900, 2013 Crossref, Medline, Google Scholar |
| 29. | Setoguchi S, Earle CC, Glynn R, et al: Comparison of prospective and retrospective indicators of the quality of end-of-life cancer care. J Clin Oncol 26:5671-5678, 2008 Link, Google Scholar |
| 30. | Weeks JC, Cook EF, O’Day SJ, et al: Relationship between cancer patients’ predictions of prognosis and their treatment preferences. JAMA 279:1709-1714, 1998 Crossref, Medline, Google Scholar |
| 31. | Mack JW, Cronin A, Keating NL, et al: Associations between end-of-life discussion characteristics and care received near death: A prospective cohort study. J Clin Oncol 30:4387-4395, 2012 Link, Google Scholar |
| 32. | Chung HO, Oczkowski SJ, Hanvey L, et al: Educational interventions to train healthcare professionals in end-of-life communication: A systematic review and meta-analysis. BMC Med Educ 16:131, 2016 Crossref, Medline, Google Scholar |
| 33. | Hui D, dos Santos R, Chisholm G, et al: Clinical signs of impending death in cancer patients. Oncologist 19:681-687, 2014 Crossref, Medline, Google Scholar |
| 34. | Hwang IC, Ahn HY, Park SM, et al: Clinical changes in terminally ill cancer patients and death within 48 h: When should we refer patients to a separate room? Support Care Cancer 21:835-840, 2013 Crossref, Medline, Google Scholar |
