Abstract
Background:
Self-reported pain is frequently reported among patients with advanced heart failure (aHF). Agreement between patients’ pain assessment and their proxies have shown inconsistent results. We examined agreement between documented patient self-report and family-reported pain scores for Veterans with aHF towards the end of life (EOL).
Methods:
We conducted a retrospective cohort study of aHF patients who died between 2018–2020 while inpatient at one of 83 Veterans Affairs Medical Centers (VAMC) and whose families completed a Beavered Family Survey (BFS) (n = 1045). BFS is a nationally distributed questionnaire assessing family-reported quality of inpatient EOL care at VAMCs. Averaged daily patient-reported numeric rating scale (NRS) pain scores over the last 30 and 7-days of life were categorized as: no pain (NRS = 0), mild (NRS = 1–3), moderate (NRS = 4–6), and severe (NRS = 7–10). NRS categories were compared to family-reported pain using responses to two BFS items (ever pain; some pain). Agreement between NRS categories and family-reported pain was assessed using Cohen’s kappa coefficient statistic.
Results:
There was moderate agreement between 30-day NRS severe pain and BFS ever pain and fair agreement for the last 7 days. Mild pain showed no-to-slight agreement compared to both BFS pain outcomes. Substantial or complete agreement was not identified between NRS and BFS. Agreement was generally stronger for last 30 vs 7 days of life.
Conclusions:
Agreement between aHF patients’ and their families’ pain reporting was moderate for the worst pain categories only and stronger with longer time for patient-reported pain. These findings suggest limitations using proxy-reported pain in patients with aHF.
Keywords: pain, numeric rating scale, bereaved family survey, concordance, proxy-reported
Introduction
Pain assessments in patients with serious, life-limiting conditions remain challenging.1,2 End of life (EOL) care often includes involvement of proxies as caregivers to deliver care to treat uncomfortable symptoms and other sources of suffering for the patient. This requires proxy symptom assessments, including spouses and adult children, to manage distressing symptoms for patients near EOL. Proxy-reported pain assessments are used to evaluate care quality, particularly palliative and hospice care delivery near the EOL.
Agreement in symptom assessment between patients and their proxies is inconsistent, although pain assessments may perform more favorably relative to other symptoms.3 Some evidence suggests greater agreement between patients and their clinical nurses than their family caregivers in hospice populations.4 However, comparisons between patients and proxies’ symptom assesments have primarily focused on cancer populations, people with dementia and geriatrics groups.5,6
Pain affects between 35%–78% of patients with advanced heart failure (aHF).7–10 In addition to ischemia-related pain, people with aHF who are older, with worse aHF functional class and with co-occuring musculoskeletal pain towards the EOL could be more likely to experience pain.11,12 Pain in aHF has primarily relied on patient self-report. In one outpatient study of people with chronic conditions that included aHF, families overestimated symptoms, including pain, relative to patient reports.13 Few studies have compared patient and proxy-reported pain among people with HF.
Determining agreement between patients and their caregivers is important for hospice and palliative providers to calibrate their internvetions in response to patient and caregiver experiences and observations. To address these knowledge gaps, we examined the agreement between self-reported pain scores of aHF decedents and their families’ rating of their pain towards the EOL.
Methods
Patient Population
We conducted a retrospective cohort study using HEART-PAL, an electronic health record-based national cohort of Veterans.14–16 We included aHF patients (identified by International Classification of Diseases 10th Revision codes and hospitalizations17) who died while inpatient at 83 VA medical centers (VAMC) between 2018–2020 and for whom families completed a Bereaved Family Survey (BFS; n = 1045). The BFS is a nationally distributed questionnaire assessing family-reported quality of inpatient EOL care at VAMCs.18,19
Outcome
Patient-reported pain measures were derived from methods described and validated elsewhere within the VA.20 Patient-reported pain scores were collected using the Numeric Pain Rating Scale (NRS) collected from the VA Corporate Data Warehouse (CDW). Pain scores were measured daily and averaged over the last 30 and 7 days of life and categorized as: no pain (NRS 0), mild (NRS 1–3), moderate (NRS 4–6), and severe (NRS 7–10).
NRS pain score averages were compared to family-reported pain based on answers to the BFS pain questions. These questions ask families if the patient had pain (or took medication for pain); and, if present, how frequently the pain made the patient uncomfortable in their last month of life. Answers were dichotomized into two variables for the agreement analysis1: Ever pain (pain present and always/usually/sometimes vs never made them uncomfortable) and2 Some pain (family-reported pain present and always/usually vs sometimes/never made them uncomfortable).
Other Clinical Variables
We collected patient demographics including age, race, ethnicity, sex, and comorbidities using the Charlson Comorbidity Index (CCI) (categorized as 0–1 and ≥2; also Charlson-derived history of cancer given association with pain and HF diagnosis separately) at the time of aHF diagnosis using CDW data. We identified palliative care receipt during the study period using validated algorithms,15 and measured the number of days from palliative consult until EOL. Finally, in addition to the BFS pain outcomes, we collected information on next of kin (spouse, adult child, sibling, parent, friend, other (Partner, Exspouse, POA/legal guardian/caretaker, Other, Did not answer)).
Statistical Analysis
We used descriptive and bivariate statistics include chi-square and analysis of variance tests to characterize demographic, clinical, and facility characteristics of the total sample and by NRS categories in the last 30 and 7 days of life. We assessed the agreement between patient-reported pain and family-reported pain measures (ever pain, some pain) using the Cohen’s kappa coefficient statistic during both time periods. We categorized the degree of agreement as poor (Kappa ≤ 0.20), fair (Kappa = 0.21–0.40), moderate (Kappa = 0.41–0.60), substantial (Kappa = 0.61–0.80) and excellent (Kappa = 0.81–0.99).
Results
Among 1045 patients with aHF, the majority reported pain during the last 30 (n = 982, 94%) and 7 days of life (n = 890, 85%). Moderate pain was most common (n = 451 (43%) and n = 443 (42%), respectively) followed by mild pain (n = 429 (41%) and n = 282 (27%)) and 10% and 16% reported severe pain (Table 1). Median time between aHF and palliative consultation was 7 days [interquartile range 1–66]; longer time to palliative consultation was observed for those with more significant pain. Of the sample, 78% of patients received care in facilities with a palliative care unit on site.
Table 1.
Decedents With Family-Completed BFS in VAMS 2018–2020 by Patient-Reported Pain (NRS) in the Last 7 and Last 30 days of Life (n = 1045)
| Demographics | Overall N = 1045 |
Pain categories for last 7 Days of life | P-value | Pain categories for last 30 days of life | P-value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| None (n = 155) |
Mild (n = 282) |
Moderate (n = 443) |
Severe (n = 165) |
None (n = 63) |
Mild (n = 429) |
Moderate (n = 451) |
Severe (n = 102) |
||||
| Mean pain (SD), last 30 days | 3.6 (2.5) | 0.9 (1.4) | 1.8 (1.3) | 4.5 (1.6) | 6.6 (1.9) | <.001 | 0 (0) | 1.5 (0.9) | 5 (1.1) | 8 (0.8) | <.001 |
| Mean pain (SD), last 7 days | 3.9 (2.9) | 0 (0) | 1.6 (0.8) | 5.1 (1.1) | 8.3 (0.8) | <.001 | 0 (0) | 2 (2) | 5.3 (2) | 7.8 (1.3) | <.001 |
| Age at aHF, years (SD) | 79 (10.2) | 82 (11.2) | 81 (9.3) | 79 (10.3) | 78 (10.1) | <.001 | 86 (8.9) | 81 (9.9) | 78 (10.1) | 78 (9.9) | <.001 |
| Race | .7 | .2 | |||||||||
| White | 770 (73.7) | 113 (72.9) | 204 (72.3) | 328 (74) | 125 (75.8) | 46 (73) | 313 (73) | 342 (75.8) | 69 (67.7) | ||
| Black | 206 (19.7) | 34 (21.9) | 54 (19.2) | 90 (20) | 28 (17) | 12 (19.1) | 86 (20.1) | 85 (18.9) | 23 (22.6) | ||
| Other | 23 (2.2) | 1 (0.7) | 10 (3.6) | 9 (2) | 3 (1.8) | 0 (0) | 13 (3) | 9 (2) | 1 (0.9) | ||
| Unknown | 46 (4.4) | 7 (4.5) | 14 (5) | 16 (3.6) | 9 (5.5) | 5 (7.9) | 17 (4) | 15 (3.3) | 9 (8.8) | ||
| Hispanic | 73 (7) | 17 (11) | 29 (10.3) | 21 (4.7) | 6 (3.6) | .02 | 10 (15.9) | 42 (9.8) | 18 (4) | 3 (2.9) | <.001 |
| Male | 1024 (98) | 154 (99.4) | 279 (98.9) | 431 (97.3) | 160 (97) | .2 | 63 (100) | 421 (98.1) | 443 (98.2) | 97 (95.1) | .1 |
| Charlson comorbidity | .04 | .01 | |||||||||
| 0–1 | 166 (15.9) | 39 (25.2) | 47 (16.7) | 59 (13.3) | 21 (12.7) | 16 (25.4) | 79 (18.4) | 55 (12.2) | 16 (15.7) | ||
| ≥2 | 879 (84.1) | 116 (74.8) | 235 (83.3) | 384 (86.7) | 144 (87.3) | 47 (74.6) | 350 (81.6) | 396 (87.8) | 86 (84.3) | ||
| Cancer | 182 (17.4) | 25 (16.1) | 49 (17.4) | 81 (18.3) | 27 (16.4) | .9 | 8 (12.7) | 72 (16.8) | 88 (19.5) | 14 (13.7) | .33 |
| HF Charlson | 745 (71.3) | 100 (64.5) | 189 (67) | 330 (74.5) | 126 (76.4) | .02 | 40 (63.5) | 291 (67.8) | 340 (75.4) | 74 (72.6) | .04 |
| Palliative care consultation | 653 (62.5) | 84 (54.2) | 177 (62.8) | 285 (64.3) | 107 (64.9) | .1 | 32 (50.8) | 268 (62.5) | 291 (64.5) | 62 (60.8) | .20 |
| Time from aHF to palliative consult, days (IQR) | 7 (1–66) | 6.5 (1–66) | 5 (1–32) | 9 (1–73) | 16 (1–114) | .1 | 4 (1–59) | 6 (1–39) | 9 (1–103) | 20 (2–79) | .06 |
| Time from PC consultation to death, days (IQR) | 9 (3.3–33.4) | 6.4 (2.2–32.4) | 6.3 (3.1–20.6) | 9.8 (3.4–34.1) | 16.8 (6.8–51.8) | .5 | 4 (2.2–9.5) | 8.3 (3.1–24.6) | 10.3 (4–36.3) | 15.1 (4.7–54.3) | .3 |
| Facility-level factors | |||||||||||
| PCU On site | 813 (77.8) | 119 (76.8) | 216 (76.6) | 353 (79.7) | 125 (75.8) | .7 | 50 (79.4) | 320 (74.6) | 363 (80.5) | 80 (78.4) | .2 |
| Facility complexity | 1.0 | .9 | |||||||||
| 0 | 594 (56.8) | 91 (58.7) | 163 (57.8) | 248 (56) | 92 (55.8) | 35 (55.6) | 250 (58.3) | 248 (55) | 61 (60) | ||
| 1 | 221 (21.2) | 30 (19.4) | 57 (20.2) | 100 (22.6) | 34 (20.6) | 15 (23.8) | 89 (20.8) | 95 (21.1) | 22 (21.6) | ||
| 2 | 230 (22) | 34 (21.9) | 62 (22) | 95 (21.4) | 39 (23.6) | 13 (20.6) | 90 (21) | 108 (24) | 19 (18.6) | ||
| VAMC geographic region | .02 | <.001 | |||||||||
| Midwest | 232 (22.2) | 24 (15.5) | 57 (20.2) | 101 (22.8) | 50 (30.3) | 7 (11.1) | 100 (23.3) | 98 (21.7) | 27 (26.5) | ||
| Northeast | 166 (15.9) | 30 (19.4) | 41 (3.9) | 68 (15.4) | 27 (16.4) | 16 (25.4) | 73 (17) | 61 (13.5) | 16 (15.7) | ||
| South | 472 (45.2) | 64 (41.3) | 142 (50.4) | 203 (45.8) | 63 (38.2) | 25 (39.7) | 190 (44.3) | 213 (47.2) | 44 (43.1) | ||
| West | 141 (13.5) | 27 (17.4) | 28 (9.9) | 63 (38.2) | 23 (13.9) | 10 (15.9) | 42 (9.8) | 74 (16.4) | 15(914.7) | ||
| Other | 34 (3.3) | 10 (6.5) | 14(5) | 8 (1.8) | 2 (1.2) | 5 (7.9) | 24 (5.6) | 5 (1.1) | 0 (0) | ||
| BFS data from next of kin | |||||||||||
| Next of kin relationship to patient | .07 | .02 | |||||||||
| Spouse | 471 (45.1) | 65 (41.9) | 132 (46.8) | 197 (44.5) | 77 (46.7) | 29 (46) | 195 (45.5) | 207 (45.9) | 40 (39.2) | ||
| Child | 309 (29.6) | 58 (37.4) | 84 (29.8) | 128 (28.9) | 39 (23.6) | 24 (38.1) | 134 (31.2) | 133 (29.5) | 18 (17.7) | ||
| Sibling | 113 (10.8) | 16 (10.3) | 20 (7.1) | 51 (11.5) | 26 (15.8) | 6 (9.5) | 34 (8.2) | 47 (10.4) | 25 (24.5) | ||
| Parent | 28 (2.7) | 1 (0.7) | 8 (2.8) | 12 (2.7) | 7 (4.2) | 0 (0) | 8 (1.9) | 16 (3.6) | 4 (3.9) | ||
| Friend | 27 (2.6) | 3 (2.9) | 7 (2.5) | 15 (3.4) | 2 (1.2) | 0 (0) | 14 (3.3) | 11 (2.4) | 2 (1.9) | ||
| Other relatives | 46 (4.4) | 6 (3.9) | 20 (7.1) | 14 (3.2) | 6 (3.6) | 0 (0) | 27 (6.3) | 13 (2.9) | 6 (5.9) | ||
| Other category (partner, ex-spouse, POA/legal guardian/caretaker, other, did not answer) | 51 (4.9) | 6 (3.9) | 11 (3.9) | 26 (5.9) | 8 (4.9) | 1 (1.6) | 12 (2.8) | 15 (3.3) | 4 (3.9) | ||
| BFS item responses | |||||||||||
| Ever pain (pain (always/usually/sometimes vs never) | 812 (77.7) | 90 (58.1) | 201 (71.3) | 375 (84.7) | 146 (88.5) | <.001 | 27 (42.9) | 301 (70.2) | 390 (86.5) | 94 (92.2) | <.001 |
| Some pain (pain always/usually vs sometimes/never | 524 (50.1) | 45 (29) | 119 (42.2) | 256 (57.8) | 104 (63) | <.001 | 11 (17.5) | 171 (39.9) | 274 (60.8) | 68 (66.7) | <.001 |
| Pain always vs other | 290 (27.8) | 25 (16.1) | 56 (19.9) | 145 (32.7) | 64 (38.8) | <.001 | 2 (3.2) | 84 (19.6) | 165 (36.6) | 39 (38.2) | <.001 |
Abbreviations: SD = standard deviation; IQR = interquartile ranges; aHF = advanced heart failure; VAMC = VA medical center; BFS = bereaved family survey; PCU = palliative care unit; POA = power of attorney.
Mean age at aHF diagnosis was 79 years (standard deviation 10.2), 98% were men and findings did not vary by NRS pain categories. Self-identified race and ethnicity were white = 74%, black = 20%, and Hispanic = 7%. Pain prevalence increased with multimorbidity measured by CCI ≥2 compared with CCI = 0 or 1 in both time periods. Cancer prevalence was similar across pain groups (P > .05).
Most commonly the BFS respondents identified as a spouse to the decedent (45%); followed by an adult child (30%) and a sibling (11%). Among patients with any pain the last 30 days of life, the proportion of family caregivers describing ever pain increased with pain category (70% among those with mild pain, 87% for those with moderate and 92% with severe pain); similar patterns were observed for those with some pain across pain categories (40%, 61% and 67%, respectively) and within the last 7 days of life.
Overall, NRS pain category increased with family-reported BFS pain. Among those with pain in the past 30 days, moderate pain was most commonly observed in family-reported ever and some pain categories (for example in ever pain = 390/812, 48% vs mild = 37%, severe = 12%; Table 1). Among patients who reported no pain (NRS = 0), BFS pain was still reported, although this was uncommon (no pain = 3%).
Agreement was moderate among those with severe pain in the last 30 days of life and family-reported measures (NRS 7–10 and Ever pain vs other; K = 0.523; NRS 7–10 and Some pain vs other; K = 0.460; Figure 1). Overall agreement for ever pain and NRS 1–10 vs NRS = 0 was K = 0.164 and for some pain K = 0.079. There was fair agreement in the severe pain categories for the last 7-days of life for both family-reported pain measures (NRS 7–10 and Ever pain vs other; K = 0.309; NRS 7–10 and Some pain vs other K = 0.339). Mild and moderate pain for the last 30 and 7 days of life showed poor-to-fair agreement with family-reported pain measures (Pain ever vs other: K = 0.162 and 0.134, respectively; Some Pain vs other: K = 0.084 and 0.112, respectively). Overall agreement for ever pain in the last 7 days of life and NRS 1–10 vs NRS = 0 was K = 0.191 and for some pain K = 0.126. None of the NRS pain categories within the last 30 or 7 days of life showed substantial or complete agreement with family-reported pain measures.
Figure 1.
Agreement Between Patient -Reported Pain Severity (Mild, Moderate, Severe) Using NRS and Family -Reported Pain Complains (Ever, Some vs. Other) on BFS During Last 30 and 7 days of Life
Discussion
In this national sample of Veteran decedents with aHF we found only moderate agreement with severe clinical pain scores and family reported pain in the last 30 days of life; otherwise, agreement was poor to fair between NRS pain categories and family-reported pain at the EOL. These data suggest that proxies could be more atuned to symptoms that are less well controlled and more severe. Surprisingly, there was stronger agreement between decedents and their proxies when asking about the final 30 rather than the last 7 days of life for patients with severe pain. This could reflect that families more accurately identify their loved one’s experiences when consider a longer historical timeframe (the last 30 vs 7 days of life). While valuing patient-reported symptoms is paramount, constructive engagement with proxies may offer additional opportunities to alleviate suffering for patients with aHF at the EOL.
Our study advances the understanding of concordance, and lack thereof, between patients, caregivers and other proxies with noncancer conditions.21 A cross-sectional Dutch study of outpatients with advance lung disease, heart failure (n = 45/159 pairs; 28%) or renal dysfunction found that families overestimated symptom prevalence and severity.13 The authors found substantial agreement in pain reporting (kappa = 0.61), although they were unable to demonstrate differences in clinical subgroups. In another study including a general cohort of 576 older adults (188 with caregivers who completed proxy assessments of patient-relevant outcomes), the kappa between patient and caregiver pain was 0.58 and proxies underestimated symptom severity at higher patient-reported levels.3
Our study also illustrates how pain prevalence and concordance between patients and families changes over time, with stronger agreement for moderate and severe pain within 30 relative to 7 days before EOL. However, agreement remained moderate at best, and only for the most severe pain (NRS 7–10). Lower agreement between patient-based NRS scores in the EHR and proxies in our study may be related to unique characteristics in our study population, including only including aHF patients who died with recent engagement in a VA system. Of note, more than half of proxies identified patients as having pain at the EOL, even among patients who were identified as having no pain by average NRS scores. This suggests that family members/proxies report greater pain prevalence and severity relative to clinical documentation.
What could account for differences in pain prevalence and severity among patients and proxies? It is likely that proxy pain reports for patients with aHF evolve following the death of a loved one. Compared with concurrent comparisons between patients and family members, bereaved family members’ recollections of patient symptoms could also be impacted by grief.22 Timing from the death can also influence proxy-reported patient symptoms, with perceived intensity of pain decreasing as more time passes.23 It is also possible that proxies are biased by impressions from one particularly distressing day rather than average scores over the last 30 days of life. Further, patient-reported pain towards the end of life, more proximate to the dying process, could actually be nurse-reported pain using behavioral pain scales rather than patient-reported for those unable to respond to questions. These issues illustrate the value of longitudinal measures for proxy symptom measurement both in real time and during bereavement.
Importantly, how health systems choose to incorporate proxy-reported pain symptoms requires critical assessment, particularly in circumstances where family-reported measures are weighed as quality metrics. While overall population-level pain measures can be informative and trended, patient pain severity and individual proxy factors including time since death and proxy emotional state could also impact reported pain.
There are several limitations to be considered. Importantly, we relied on NRS asked of patients and recorded by nurses during patients’ last 30 and 7 days of life, rather than patiententered pain reports. In some instances, patients may have been unable to report pain with nurses relying on objective assessments for this documentation. It is also possible the measurements do not have adequate discriminant capacity statistically. Further, we were unable to enhance patient-reported pain with items such as facial expressions due to relying on measures already available in the EHR. We also compared pain reports between patients and families/proxies asynchronously relative to when the pain might have been experienced, with BFS reported several weeks after the patient died. We also did not have context for family-rated pain nor were we able to explore cultural differences in how patients and families perceive and discuss pain experiences. Further, we did not identify the source of pain, whether somatic or psychological or some other form of suffering. We also were unable to determine how pain was responded to by the treatment team. Finally, nearly all of the aHF decedents were men, limiting our ability to evaluate how sex or gender identify may influence correlations between NRS and family-reported pain.
Conclusions
We found moderate agreement between aHF patients’ and their families’ pain reporting for severe pain. Agreement was minimal for mild or moderate pain between NRS and family-reported pain measures. This could reflect that severe pain was more noticeably uncomfortable, and perhaps more distressing to families than other pain categories. These findings suggest limitations in using proxy-reported pain in patients with aHF. Our study findings suggest exercising caution in using proxy-reported pain measures to assess EOL care quality.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Veteran’s Health Administration Health services & Research; CIN 13-407, LIPS 96-062. This work was supported by the Department of Veterans Affairs, Veterans Health Administration, the Pain Research, Informatics, Multiple Morbidities and Education Center of Excellence (West Haven, CT) and by National Heart, Lung, and Blood Institute, which had no role in the design, methods, participant recruitment, data collection, analysis, or preparation of this article or in the decision to submit this article for publication.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Disclaimer
The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the United States Department of Veterans Affairs or the United States Government.
References
- 1.Sebring K, Shattuck J, Berk J, Boersma I, Sillau S, Kluger BM. Assessing the validity of proxy caregiver reporting for potential palliative care outcome measures in parkinson’s disease. Palliat Med. 2018;32(9):1522–1528. doi: 10.1177/0269216318785830 [DOI] [PubMed] [Google Scholar]
- 2.McPherson CJ, Addington-Hall JM. Judging the quality of care at the end of life: can proxies provide reliable information? Soc Sci Med. 2003;56(1):95–109. doi: 10.1016/s0277-9536(02)00011-4 [DOI] [PubMed] [Google Scholar]
- 3.Kroenke K, Stump TE, Monahan PO. Agreement between older adult patient and caregiver proxy symptom reports. J Patient-Rep Outcomes. 2022;6(1):50. doi: 10.1186/s41687-022-00457-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kutner JS, Bryant LL, Beaty BL, Fairclough DL. Symptom distress and quality-of-life assessment at the end of life: the role of proxy response. J Pain Symptom Manag. 2006;32(4): 300–310. doi: 10.1016/j.jpainsymman.2006.05.009 [DOI] [PubMed] [Google Scholar]
- 5.Jones JM, McPherson CJ, Zimmermann C, Rodin G, Le LW, Cohen SR. Assessing agreement between terminally ill cancer patients’ reports of their quality of life and family caregiver and palliative care physician proxy ratings. J Pain Symptom Manag. 2011;42(3):354–365. doi: 10.1016/j.jpainsymman.2010.11.018 [DOI] [PubMed] [Google Scholar]
- 6.Roydhouse JK, Cohen ML, Eshoj HR, et al. The use of proxies and proxy-reported measures: a report of the international society for quality of life research (ISOQOL) proxy task force. Qual Life Res. 2022;31(2):317–327. doi: 10.1007/s11136-021-02937-8 [DOI] [PubMed] [Google Scholar]
- 7.Walke LM, Byers AL, Tinetti ME, Dubin JA, McCorkle R, Fried TR. Range and severity of symptoms over time among older adults with chronic obstructive pulmonary disease and heart failure. Arch Intern Med. 2007;167(22):2503–2508. doi: 10.1001/archinte.167.22.2503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Godfrey CM, Harrison MB, Friedberg E, Medves JM, Tranmer JE. The symptom of pain in individuals recently hospitalized for heart failure. J Cardiovasc Nurs. 2007;22(5):368–374. doi: 10.1097/01.JCN.0000287035.77444.d9 [DOI] [PubMed] [Google Scholar]
- 9.Goebel JR, Doering LV, Evangelista LS, et al. A comparative study of pain in heart failure and non-heart failure veterans. J Card Fail. 2009;15(1):24–30. doi: 10.1016/j.cardfail.2008.09.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Goodlin SJ, Wingate S, Albert NM, et al. Investigating pain in heart failure patients: the pain assessment, incidence, and nature in heart failure (PAIN-HF) study. J Card Fail. 2012;18(10): 776–783. doi: 10.1016/j.cardfail.2012.07.007 [DOI] [PubMed] [Google Scholar]
- 11.Warraich HJ, Wolf SP, Troy J, et al. Differences between patients with cardiovascular disease and cancer referred for palliative care. Am Heart J. 2021;233:5–9. doi: 10.1016/j.ahj.2020.11.016 [DOI] [PubMed] [Google Scholar]
- 12.Levenson JW, McCarthy EP, Lynn J, Davis RB, Phillips RS. The last six months of life for patients with congestive heart failure. J Am Geriatr Soc. 2000;48(S1):S101–S109. doi: 10.1111/j.1532-5415.2000.tb03119.x [DOI] [PubMed] [Google Scholar]
- 13.Janssen DJ, Spruit MA, Wouters EF, Schols JM. Symptom distress in advanced chronic organ failure: disagreement among patients and family caregivers. J Palliat Med. 2012;15(4): 447–456. doi: 10.1089/jpm.2011.0394 [DOI] [PubMed] [Google Scholar]
- 14.Feder SL, Han L, Zhan Y, et al. Variation in specialist palliative care reach and associated factors among people with advanced heart failure in the department of veterans affairs. J Pain Symptom Manag. 2024;68(1):22–31. doi: 10.1016/j.jpainsymman.2024.03.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Feder SL, Zhan Y, Abel EA, et al. Validation of electronic health record-based algorithms to identify specialist palliative care within the department of veterans affairs. J Pain Symptom Manag. 2023;66(4):e475–e483. doi: 10.1016/j.jpainsymman.2023.06.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Feder SL, Murphy TE, Abel EA, et al. Incidence and trends in the use of palliative care among patients with reduced, middlerange, and preserved ejection fraction heart failure. J Palliat Med. 2022;25(12):1774–1781. doi: 10.1089/jpm.2022.0093 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dunlay SM, Blecker S, Schulte PJ, Redfield MM, Ngufor CG, Glasgow A. Identifying patients with advanced heart failure using administrative data. Mayo Clin Proc Innov Qual Outcomes. 2022;6(2):148–155. doi: 10.1016/j.mayocpiqo.2022.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Thorpe JM, Smith D, Kuzla N, Scott L, Ersek M. Does mode of survey administration matter? Using measurement invariance to validate the mail and telephone versions of the bereaved family survey. J Pain Symptom Manag. 2016;51(3):546–556. doi: 10.1016/j.jpainsymman.2015.11.006 [DOI] [PubMed] [Google Scholar]
- 19.Casarett D, Shreve S, Luhrs C, Lorenz K, Smith D, De Sousa M, Richardson D. Measuring families’ perceptions of care across a health care system: preliminary experience with the family assessment of treatment at end of life short form (FATE-S). J Pain Symptom Manag. 2010;40(6):801–809. doi: 10.1016/j.jpainsymman.2010.03.019 [DOI] [PubMed] [Google Scholar]
- 20.Goulet JL, Brandt C, Crystal S, et al. Agreement between electronic medical record-based and self-administered pain numeric rating scale: clinical and research implications. Med Care. 2013; 51(3):245–250. doi: 10.1097/MLR.0b013e318277f1ad [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hack TF, McClement SE, Chochinov HM, et al. Assessing symptoms, concerns, and quality of life in noncancer patients at end of life: how concordant are patients and family proxy members? J Pain Symptom Manag. 2018;56(5):760–766. doi: 10.1016/j.jpainsymman.2018.07.019 [DOI] [PubMed] [Google Scholar]
- 22.Higginson I, Priest P, McCarthy M. Are bereaved family members a valid proxy for a patient’s assessment of dying? Soc Sci Med. 1994;38(4):553–557. doi: 10.1016/0277-9536(94)90251-8 [DOI] [PubMed] [Google Scholar]
- 23.McPherson CJ, Addington-Hall JM. How do proxies’ perceptions of patients’ pain, anxiety, and depression change during the bereavement period? J Palliat Care. 2004;20(1):12–19. [PubMed] [Google Scholar]