Abstract
BACKGROUND/OBJECTIVES:
To assess whether an advance care planning (ACP) video intervention impacts care among short-stay nursing home (NH) patients.
DESIGN:
PRagmatic trial of Video Education in Nursing Homes (PROVEN) was a pragmatic cluster randomized clinical trial.
SETTING:
A total of 360 NHs (N = 119 intervention, N = 241 control) owned by two healthcare systems.
PARTICIPANTS:
A total of 2,538 and 5,290 short-stay patients with advanced dementia or cardiopulmonary disease (advanced illness) in the intervention and control arms, respectively; 23,302 and 50,815 short-stay patients without advanced illness in the intervention and control arms, respectively.
INTERVENTION:
Five ACP videos were available on tablets or online. Designated champions at each intervention facility were instructed to offer a video to patients (or proxies) on admission. Control facilities used usual ACP practices.
MEASUREMENTS:
Follow-up time was at most 100 days for each patient. Outcomes included hospital transfers per 1000 person-days alive and the proportion of patients experiencing more than one hospital transfer, more than one burdensome treatment (tube-feeding, parenteral therapy, invasive mechanical intervention, and intensive care unit admission), and hospice enrollment. Champions recorded whether a video was offered in the patients’ electronic medical record.
RESULTS:
There was no significant reduction in hospital transfers per 1000 person-days alive in the intervention versus control groups with advanced illness (rate (95% confidence interval (CI)), 12.3 (11.6–13.1) vs 13.2 (12.5–13.7); rate difference: −0.8; 95% CI = −1.8–0.2)). There was a nonsignificant reduction in hospital transfers per 1000 person-days alive in the intervention versus control among short-stay patients without advanced illness. Secondary outcomes did not differ between groups among patients with and without advanced illness. Based on champion only reports 14.2% and 15.3% of eligible short-stay patients with and without advanced illness were shown videos, respectively.
CONCLUSION:
An ACP video program did not significantly reduce hospital transfers, burdensome treatment, or hospice enrollment among short-stay NH patients; however, fidelity to the intervention was low.
Keywords: pragmatic clinical trial, nursing homes, advance care planning, Medicare
INTRODUCTION
In 2018, nursing homes (NHs) provided rehabilitation services to 1.6 million fee-for-service (FFS) Medicare beneficiaries.1 Most (40%)1 are destined to return home following rehabilitation but, for some, particularly those with advanced illness, postacute care is the beginning of a terminal decline and death.2,3 Burdensome transitions at end of life such as rehospitalization and dying in the hospital are common.4–7 High rates of hospitalization coupled with low hospice selection among short-stay NH decedents7,8 contributes to low-quality end-of-life care.
One modifiable factor that may contribute to poor quality care at end of life is the lack of optimal advance care planning (ACP).9–12 ACP is the process of communicating healthcare preferences in the event that the patient is unable to communicate their wishes. Although the Centers for Medicare and Medicaid Services require NHs to ascertain preferences upon admission, practices are inconsistent.13 Video ACP tools, which were previously tested in efficacy trials, were developed to improve communication of complex health information about preferences.14–21
PRagmatic trial Of Video Education in NHs (PROVEN) was a pragmatic cluster randomized control trial designed to test the effectiveness of ACP videos on the care of NH patients. Although the population for the trial was all patients in the NH, the primary trial outcome was specified for a target population of long-stay residents with advanced illness. The main trial results are reported elsewhere.22 Briefly, the study found that the ACP program did not have statistically significant effect on hospital transfers, burdensome treatment use, or hospice enrollment among long-stay residents with or without advanced illness. The intervention fidelity among long-stay residents was also low and variable across facilities. The trial protocol also prespecified the examination of outcomes for patients admitted to NHs for shorter duration (<100 days). This report presents the following outcomes in this short-stay population measured up to 100 days of follow-up: hospital transfers, use of burdensome interventions, and hospice enrollment.
METHODS
Brown University’s Institutional Review Board approved the study with a waiver of informed consent (Clinical Trial Registration: NCT02612688). Details of the trial protocol are reported elsewhere.23
Randomization
The PROVEN trial was conducted between December 2015 and May 2019 in two large for-profit healthcare systems. There were 360 NHs, 119 interventions, and 241 controls. NHs were eligible to be randomized if they met the following criteria: over 50 beds and both long- and short-stay patients. Among eligible facilities, NH leadership excluded those with organizational problems (e.g., being a Special Focus Facility) or inability to transfer electronic health records.
NHs were stratified by healthcare system and then by the proportion of hospitalizations per 1000 person-days alive among residents with advanced illness using 2013 minimum data set (MDS).24 Within each strata, NHs were randomized to intervention and control with a 1:2 ratio.
Participants
Participant enrollment began March 1, 2016, and ended May 31, 2018, with 100-day follow-up for each short-stay patient. Patients eligible for analysis in this report were: (1) aged 65 and older, (2) length of stay less than 100 days in the NH, and (3) Medicare FFS. The protocol dictated that everyone admitted to the intervention facilities during enrollment should be offered to watch a video within the first week of admission.
Prespecified outcomes are reported for short-stay patients meeting the aforementioned eligibility criteria with and without advanced illness. Advanced illness was defined as having either advanced dementia or cardiopulmonary disease (chronic obstructive pulmonary disorder (COPD) or congestive heart failure (CHF)) according to the mandated MDS assessment completed during the first 100 days of their NH stay. Advanced dementia was defined as having either Alzheimer’s disease or other dementia, advanced cognitive impairment (score of 3 or 4 on the Cognitive Function Scale25) and needing extensive or total assistance for eating and transferring. Advanced cardiopulmonary disease was defined as having either COPD or CHF plus shortness of breath while sitting or supine, and needing extensive or total assistance with walking, transferring, locomotion, or dressing. Any short-stay patient who did not meet the above criteria was considered without advanced illness. Specifically, short-stay patients without advanced illness had neither advanced dementia nor advanced CHF/chronic obstructive lung disease.
Intervention
The intervention consisted of five previously created18 short videos offered in English or Spanish: (1) general goals of care; (2) goals of care for advanced dementia; (3) hospice; (4) hospitalization; and (5) ACP for health patients. Each facility was given two tablets preloaded with the videos. Each NH designated one or two ACP video program champions (mostly social workers) were charged with implementing the intervention. Champions were trained either in-person or via webinar prior to participant enrollment. The protocol instructed the champions to offer the videos to patients (or their proxies) at the following time points: (1) within 7 days of admission or readmission, (2) when decisions arose for which there was a specific video (e.g., hospice), and (3) special circumstances (e.g., major health status change). Champions were instructed that the video was to supplement, not replace, existing ACP practices. Training included guidance on how to initiate ACP conversations and integrate the videos into the discussions. The primary mode of viewing was on the tablet at the NH. Patients or proxies could decline to watch the video once offered. A password-protected link was also made available to proxies who preferred to view the video at another time or to those who could not come to the NH. Champions were instructed to record whenever a video was offered and whether the video was watched by the patient (or proxy) in a video status report that was embedded into all facilities’ electronic medical records. If a video was not shown, the champion recorded the reason why as of the following options (check all that applied): video links given to family or nonfamily responsible party, patient refused, family or nonfamily responsible party refuse, patient was medically unstable, staff feels video is not indicated at this time, and other. Each month, the research team provided feedback reports of the facilities offer and show rate. Control NHs used their usual ACP procedures.
Data Sources and Baseline Variables
Data were derived from Medicare claims, MDS assessments, and video status reports. We ascertained baseline patient characteristics from the admission MDS assessments. Demographic characteristics included: age, gender, race/ethnicity (white, black, Asian, Hispanic, Other: Native Hawaiian/Other Pacific Islander, American Indian/Alaska Native, or more than one race, Unknown), NH length of stay and follow-up time. Functional status was quantified using the MDS Activities of Daily Living Scale (range 0–28 with higher scores indicating greater functional dependence).26 The patients Mortality Risk Score27 (MRS) (range 0–39, higher score indicates greater risk of death) was also calculated from MDS data. Hospice enrollment was determined using Medicare claims.
Outcomes
The main outcome was the number of hospital transfers per 1000 person-days alive. We followed each patient for up to 100 days, starting from NH admission or the first day the study began. Death dates were determined using the Medicare enrollment file. If patients switched to Medicare Advantage during the 100 days, their last day of follow-up was the last day of Medicare FFS coverage. We treated hospital transfers as mutually exclusive in the following categories as measured using Medicare claims: inpatient transfer, emergency department, or observation stay.
We conducted analysis of the following additional binary outcomes measured as the proportion of patients who experienced the following in 100 days post NH admission: at least one hospital transfer, at least one burdensome intervention, and hospice enrollment. Burdensome interventions were determined from MDS assessments and Medicare claims and included: tube-feeding, parenteral therapy, invasive mechanical intervention, and admission to an intensive care unit.28–33 We measured intervention fidelity as the proportion of patients offered and shown a video at least once based on champion documented video status reports.
Statistical Analysis
We described baseline characteristics using means and standard deviations (SDs) and proportions for continuous and categorical variables, respectively. Analysis followed the intention-to-treat principle. To test the effect of the intervention on hospital transfers per 1000 person-days alive, we used a hierarchical zero-inflated Poisson model to account for facility clustering with an indicator for intervention facility.34 We used a two-sided test of the difference in marginal means to examine the null hypothesis. We calculated standard errors (SEs) using the bootstrap procedure.35 We used hierarchical logistic regression to estimate the effect of the intervention on binary secondary outcomes. We excluded patients on hospice at baseline from the hospice analysis. Based on the logistic regression model, we estimated the marginal differences in proportions and 95% confidence intervals (CIs) were reported for these analyses.
Because the group of short-stay patients without our target advanced illness is heterogeneous and had a large sample size, we conducted post-hoc analyses categorizing this group by terciles of MRS scores (low: 1–3; Medium: 4–5; High: 6+).27 We hypothesized that the patients without advanced illness who were in group with the high MRS category would respond to the intervention similarly to the advanced illness cohort.
RESULTS
Facility and Patient Characteristics
As shown in Figure 1, among the 454 NHs that met initial trial eligibility (having 50 or more beds and serving both long- and short-stay patients), we excluded 94 (21%) due either to organization problems (N = 82; 18%) or inability to transmit electronic records (N = 12; 3%). We randomly assigned the 360 (79%) remaining facilities in a 1:2 ratio to the intervention and control groups. There were 119 intervention facilities and 241 control facilities. One healthcare system sold one randomized facility prior to intervention. We excluded 19,534 (12%) patients that could not be matched to claims and 40,249 (25%) that were not Medicare FFS. The final analytic patient sample included 25,840 short-stay patients in the intervention arm and 55,475 in the control arm.
Figure 1.
Consort diagram.
Of the 25,840 patients in the intervention arm, 2,538 (10%) met the criteria for advanced illness and 23,302 (90%) did not have advanced illness (Table 1). The intervention and control arms were similar across many characteristics. At baseline, among those with advanced illness 102 (4.0%) of the intervention group and 227 (4.3%) of the control group were on hospice whereas only 408 (1.8%) of the intervention group and 1,026 (2.0%) of the control group were on hospice among those without advanced illness. During the 100-day follow-up period, a higher proportion of patients with advanced illness died compared to those without advanced illness, with similar distribution of mortality rates between the intervention and control arm in both cohorts (advanced illness, intervention arm, 1,106 (43.6%) died and control arm, 2,254 (42.6%) died; nonadvanced illness, intervention arm, 4,214 (18.1%) died and control arm, 8,781 (17.5%) died). Among those with advanced illness, NH average length of stay was 31.1 days (SD: 25.4) in the intervention group and 31.2 days (SD: 25.4) in the control group. NH average length of stay among those without advanced illness, was 35.4 days (SD 31.2) in the intervention group and 34.3 days (SD 30.4) in the control group.
Table 1.
Characteristics of Short-Staya Nursing Home Patients
| Characteristics | With Advanced Illnessb |
Without Advanced Illness Intervention |
||
|---|---|---|---|---|
| Intervention |
Control |
Intervention |
Control |
|
| N = 2,538 | N = 5,290 | N = 23,302 | N = 50,815 | |
| Age, mean (SD) | 82.2 (8.6) | 82.2 (8.5) | 80.0 (8.7) | 80.1 (8.7) |
| Female, n (%) | 1,477 (58.2) | 3,034 (57.3) | 13,370 (57.4) | 29,145 (58.1) |
| Race, n (%) | ||||
| White | 2,035 (80.2) | 4,413 (83.4) | 19,139 (82.1) | 42,550 (84.8) |
| Black | 416 (16.4) | 698 (13.2) | 3246 (13.9) | 5715 (11.4) |
| Asian | 18 (0.7) | 41 (0.8) | 169 (0.7) | 463 (0.9) |
| Hispanic | 24 (0.95) | 38 (0.7) | 200 (0.9) | 434 (0.9) |
| Otherc | 45 (1.8) | 101 (2.0) | 4,138 (2.3) | 1,023 (2.0) |
| Advanced illness | ||||
| Advanced dementia, n (%) | 1,350 (49.3) | 2,545 (48.1) | – | – |
| Advanced CHF/COPDb, n (%) | 1,358 (53.5) | 2,881 (54.5) | – | – |
| Hospice at baseline, n (%) | 102 (4.0) | 227 (4.3) | 408 (1.8) | 1,026 (2.0) |
| ADL scored, mean (SD) | 21.1 (3.3) | 21.0 (3.2) | 15.2 (6.4) | 15.4 (6.2) |
| MRS scoree, mean (SD) | 7.5 (2.8) | 7.4 (2.9) | 4.6 (2.2) | 4.6 (2.1) |
| Follow-up characteristics | ||||
| Died during follow-up, n (%) | 1,106 (43.6) | 2,254 (42.6) | 4,214 (18.1) | 8,781 (17.5) |
| Discharged Home | 1,060 (41.8) | 2,263 (42.8) | 14,129 (60.6) | 30,943 (61.7) |
| NH Length of stay, mean (SD) | 31.1 (25.4) | 31.2 (25.4) | 35.4 (31.2) | 34.3 (30.4) |
| Days of follow-upc, mean (SD) | 72.4 (36.3) | 73.2 (36.1) | 89.4 (25.6) | 89.8 (26.3) |
| Intervention fidelity | ||||
| Offered, n (%) | 1,583 (62.4) | – | 16,208 (69.6) | – |
| Shown, n (%) | 361 (14.2) | – | 3,572 (15.3) | – |
Abbreviations: ADL, activities of daily living; SD, standard deviation.
Short Stay <100 days in nursing home.
Advanced illness includes residents with advanced dementia, advanced congestive heart failure (CHF), and advanced chronic obstructive pulmonary disease (COPD).
Other: Native Hawaiian/Other Pacific Islander, American Indian/Alaska Native, or more than one race, Unknown.
Activities of daily living score, range 0–28, is the sum of scores in seven domains of function: bed mobility, dressing, toileting, transferring, eating, and locomotion.Each is scored on a 5-point scale (0, independent; 1, supervision; 2, limited assistance; 3, extensive assistance; and 4, total dependence.
MRS3: MDS 3.0 Mortality Risk Score, range 0–39, higher scores indicate higher risk of mortality.
Intervention Fidelity
Among short-stay patients with advanced illness in intervention facilities, 62.4% (N = 1,583) were offered and 14.2% (N = 361) were shown a video. Among short-stay patients without advanced illness, 69.6% (N = 16,208) were offered the video and 15.3% (N = 3,572) were shown a video. The most common reasons for not showing a video among those with advanced illness was patient refusal (N = 375; 30.7%), family refusal (N = 200; 16.4%), or staff felt the video was not indicated (N = 143; 11.2%). Among patients without advanced illness, the most common reasons were patient refusal (N = 6,128; 48.5%), family refusal (N = 1,112; 8.8%), and staff felt the video is not indicated (N = 1,213; 9.6%).
At the facility level, the proportion of patients actually shown a video varied substantially across NHs; 9 intervention facilities (7.5%) apparently never showed videos to the short-stay population, 50 facilities (42.0%) had shown-rates lower than 11% and 19 (16.0%) of facilities had shown-rates greater than 40%.
Outcomes
The marginal rate of hospital transfer per 1000 person-days alive for short-stay patients with advanced illness in the intervention (12.3; 95% CI = 11.6–13.1) was lower than the control group (13.2; 95% CI = 12.5–13.7) but did not differ significantly (Marginal rate differences: −0.8; 95% CI = −1.7 to 0.2) (Table 2). For the short-stay patients without advanced illness; the marginal rate of hospital transfer/1000 person-days alive in the intervention (9.5; 95% CI = 9.3–9.8) was higher than the control group (9.3; 95% CI = 9.1–9.5) but did not differ significantly (Marginal rates difference: 0.2; 95% CI = −0.1 to 0.6).
Table 2.
Outcomes of Advance Care Planning Video Intervention among Short-Stay Nursing Home Patients
| Primary Outcome |
With Advanced Illness |
Without Advanced Illness |
||||
|
Ratea (95% CI) |
Ratea (95% CI) |
|||||
| Intervention, N = 2,538 | Control, N = 5,290 | Rate Differencea, (95% CI) | Intervention, N = 23,302 | Control, N = 50,815 | Rate Differencea, (95% CI) | |
| Hospital transfersb/1000 person-days alive | 12.3 (11.6–13.1) | 13.2 (12.5–13.7) | −0.8f (−1.7 to 0.2) | 9.5 (9.3–9.8) | 9.3 (9.1–9.5) | 0.2 (−0.1–0.6) |
| Secondary Outcomesc | Proportiona (95% CI) | Proportion Differencea (95% CI) | Proportiona (95% CI) | Proportion Differencea (95% CI) | ||
| ≥1 hospital transferc | 51.3 (48.9–54.0) | 51.2 (49.6–52.9) | 0.1 (−2.9–3.2) | 46.1 (44.6–47.6) | 45.7 (44.9–46.5) | 0.4 (−1.1–1.9) |
| ≥1 burdensome interventiond | 10.1 (8.1–12.3) | 10.1 (9.0–11.3) | 0.0 (−2.2 to 2.5) | 6.0 (5.3–6.8) | 5.3 (4.8–5.8) | 0.7 (−0.2 to 1.6) |
| Enrolled in hospicee | 23.7 (21.6–25.8) | 23.4 (21.7–25.2) | 0.4 (−2.3–3.0) | 8.6 (8.1–9.2) | 8.7 (8.0–9.4) | 0.1 (−0.9–1.0) |
Abbreviation: CI, confidence interval.
Adjusted for clustering at the nursing home level using random effects.
Hospital transfers include admissions, emergency room visits and observation stays.
All secondary outcomes are measured as proportion of residents who experienced the outcome over 100 days.
Burdensome interventions include tube-feeding, parenteral therapy (for hydration or medication delivery), invasive mechanical intervention, and admission to an intensive care unit.
Residents enrolled in hospice at baseline are excluded in these analysis (N = 1,763: 329 advanced illness, 1,434, without advanced illness).
Due to rounding the difference between the intervention and control hospital transfer outcomes are not equal.
Among short-stay patients with and without advanced illness the proportions of patients who experienced any hospital transfers, burdensome interventions, or who were referred to hospice did not differ significantly between the two arms (Table 2).
Supplementary Analysis
Of the 74,117 patients without advanced illness, we classified 19,482 (26%) into the low MRS group (MRS 1–3), 19,588 (26%) in the medium MRS group (MRS 4–5), and 15,691 (21%) in the high MRS group (6+) (Table 3). We were unable to calculate the MRS score for 18,186 (24.5%) short-stay patients without advanced illness because they were discharged before a complete assessment had been completed. During the 100-day follow-up period, 1,369 (26.0%) in the intervention arm and 2,691 (25.8%) in the control arm patients in the high MRS group died. Among the low MRS group, 506 (8.0%) and 1,053 (7.7%) in the intervention and control arm, respectively, died within the 100-day follow-up.
Table 3.
Characteristics of Short-Staya Nursing Home Residents without Advanced Illnesse by Mortality Risk Score
| Characteristics | Low MRSd (MRS 1–3) |
Medium MRSd (MRS 4–5) |
High MRSd (MRS 6+) |
|||
|---|---|---|---|---|---|---|
| Intervention N = 6,303 | Control N = 13,719 | Intervention N = 6,336 | Control N = 13,252 | Intervention N = 5,264 | Control N = 10,427 | |
| Age, mean (SD) | 74.9 (5.8) | 74.9 (5.7) | 82.5 (8.8) | 82.8 (8.8) | 83.9 (8.2) | 83.7 (8.3) |
| Female, n (%) | 3,635 (57.7) | 7,923 (57.8) | 3,742 (59.1) | 8,076 (60.9) | 3,040 (57.8) | 6,026 (57.8) |
| Race, n (%) | ||||||
| White | 5,090 (80.8) | 11,441 (83.4) | 5,399 (85.2) | 11,604 (87.6) | 4,945 (87.3) | 9,177 (88.0) |
| Black | 922 (14.6) | 1601 (11.7) | 697 (11.0) | 1,167 (8.8) | 498 (9.5) | 920 (8.8) |
| Asian | 46 (0.7) | 146 (1.1) | 42 (0.7) | 136 (1.0) | 37 (0.7) | 81 (0.8) |
| Hispanic | 55 (0.9) | 141 (1.0) | 56 (0.9) | 111 (0.8) | 35 (0.7) | 73 (0.7) |
| Otherb | 190 (1.6) | 390 (2.8) | 142 (1.1) | 234 (1.8) | 99 (1.9) | 176 (1.7) |
| Hospice at baseline, n (%) | 24 (0.4) | 62 (0.5) | 33 (0.5) | 85 (0.6) | 59 (1.1) | 147 (1.4) |
| ADL scorec, mean (SD) | 13.0 (6.4) | 13.4 (6.3) | 14.9 (6.2) | 15.2 (5.9) | 16.5 (6.2) | 16.8 (6.0) |
| MRS score, mean (SD) | 2.6 (0.5) | 2.6 (0.5) | 4.4 (0.5) | 4.4 (0.5) | 7.4 (1.6) | 7.4 (1.6) |
| Follow-up characteristics | ||||||
| Died during follow-up, n (%) | 506 (8.0) | 1,053 (7.7) | 913 (14.4) | 1,820 (13.7) | 1,369 (26.0) | 2,691 (25.8) |
| Discharged Home | 4,409 (70.0) | 9,715 (70.8) | 4,032 (63.6) | 8,572 (64.7) | 2,927 (55.6) | 5,718 (54.8) |
| NH Length of stay, mean (SD) | 35.9 (30.5) | 34.2 (29.6) | 39.4 (31.1) | 37.7 (30.6) | 37.6 (30.9) | 36.8 (30.3) |
| Days of follow-up, mean (SD) | 96.0 (17.1) | 96.3 (16.9) | 92.8 (21.5) | 93.0 (21.4) | 85.0 (29.7) | 85.5 (29.2) |
| Intervention fidelity | ||||||
| Offered, n (%) | 4,852 (77.0) | – | 4,984 (78.7) | – | 4,001 (76.0) | – |
| Shown, n (%) | 962 (15.3) | – | 1,014 (16.0) | – | 786 (14.9) | – |
Abbreviations: ADL, activities of daily living; SD, standard deviation.
Short Stay <100 days in nursing home.
Other: Native Hawaiian/Other Pacific Islander, American Indian/Alaska Native, or more than one race, Unknown.
Activities of daily living score, range 0–28, is the sum of scores in seven domains of function: bed mobility, dressing, toileting, transferring, eating, and locomotion, and hygiene. Each is scored on a 5-point scale (0, independent; 1, supervision; 2, limited assistance; 3, extensive assistance; and 4, total dependence.
MRS3: MDS 3.0 Mortality Risk Score, range 0–39, higher scores indicate higher risk of mortality.
We were unable to calculate the MRS score for 18,186 (24.5%) short-stay without advanced illness.
Across all strata, there was no significant difference in the proportion of hospital transfers, burdensome interventions, or hospice referrals between intervention and control facilities (Table 4). As MRS increased, hospital transfers and the proportion of burdensome treatments and hospice increased.
Table 4.
Outcomes of Advance Care Planning Video Intervention among Short-Stay Nursing Home Residents without Advanced Illness by Mortality Risk Score
| Primary Outcome |
Low MRS (MRS 1–3) |
Medium MRS (MRS 4–5) |
High MRS (MRS 6+) | ||||||
|
Ratea (95% CI) |
Ratea (95% CI) |
Ratea (95% CI) |
|||||||
| Intervention, N = 6,303 | Control, N = 13,719 | Rate Differencea, (95% CI) | Intervention, N = 6,336 | Control, N = 13,252 | Rate Differencea, (95% CI) | Intervention, N = 5,264 | Control, N = 10,427 | Rate Differencea, (95% CI) | |
| Hospital transfersb/1000 person-days alive | 8.0 (7.7–8.4) | 7.8 (7.5–8.1) | 0.2 (−0.2 to 0.7) | 8.8 (8.4–9.2) | 9.0 (8.7–9.3) | −0.2 (−0.7–0.3) | 11.6 (11.2–12.2) | 11.4 (11.1–11.8) | 0.2 (−0.4 to 0.9) |
| Secondary Outcomesd | Proportiona (95% CI) | Proportion Differencea (95% CI) | Proportiona (95% CI) | Proportion Differencea (95% CI) | Proportiona (95% CI) | Proportion Differencea (95% CI) | |||
| ≥1 hospital transferc | 42.4 (40.7–43.9) | 40.7 (39.5–41.8) | 1.7 (−0.3 to 3.6) | 45.5 (43.6–47.3) | 46.4 (45.3–47.6) | −1.0 (−3.1 to 1.3) | 52.8 (51.0–54.9) | 53.5 (52.4–54.7) | −0.7 (−3.0–1.7) |
| ≥1 burdensome interventiond | 4.1 (3.4–4.9) | 3.3 (2.8–3.8) | 0.8 (−0.04 to 1.7) | 4.8 (3.9–5.7) | 4.9 (4.3–5.6) | −0.1 (−1.2 to 1.0) | 8.6 (7.3–10.0) | 8.3 (7.4–9.4) | 0.3 (−1.4 to 1.9) |
| Enrolled in hospicee | 3.2 (2.6–3.8) | 3.1 (2.7–3.5) | 0.2 (−0.5 to 0.9) | 7.8 (6.7–8.8) | 7.1 (6.4–7.9) | 0.7 (−0.6 to 2.0) | 13.3 (12.1–14.6) | 13.3 (12.3–14.4) | 0.0 (−1.7 to 1.47) |
Abbreviation: CI, confidence interval.
Adjusted for clustering at the nursing home level using random effects.
Hospital transfers include admissions, emergency room visits and observation stays.
All secondary outcomes are measured as proportion of residents who experienced the outcome over 100 days.
Burdensome interventions include tube-feeding, parenteral therapy (for hydration or medication delivery), invasive mechanical intervention, and admission to an intensive care unit.
Residents enrolled in hospice at baseline are excluded in these analysis (N = 1,434).
DISCUSSION
We implemented a large embedded pragmatic cluster clinical trial in 360 NHs and found that fidelity was, on average low, and highly variable, which may explain why we found no statistically significant differences in in the proportion of short-stay patients that experienced hospital transfers, burdensome transitions, or enrolled in hospice.
This study’s finding was similar to the main PROVEN trial’s results with long-stay residents.22 Although our video status report data showed higher offer rates among the short-stay population (62.4%) compared to the long-stay population (55.6%) the show-rate was lower in the short-stay population (14.2% vs 21.9%).22 This low level of intervention fidelity speaks to the difficulty of implementing pragmatic trials in healthcare settings with constrained resources and competing priorities. Early reports from PROVEN found NHs with a Nursing Home Compare 5-star rating and those with greater champion engagement (measured by attendance on monthly phone calls) had higher intervention fidelity.36 Study investigators and corporate leadership assumed that staff could and would incorporate offering the video into the regular admission process. However, competing priorities meant that champions did not readily incorporate showing a short video into NHs’ standard admission orientation process. Earlier qualitative reports from PROVEN, found ACP champions, in NHs with low intervention fidelity noted that they had limited time to devote to offering and showing the video, particularly in facilities with high staff turnover.37 Some champions suggested that they would prefer to use the ACP videos on an as-needed basis to their discretion.38 Insufficient staff time to implement a new treatment or intervention is not unique to ACP. Similar difficulties were reported in a clinical intervention pilot study designed to improve depression care for short-stay NH patients which found that it was difficult to incorporate families into care planning meetings and most short stay patients had only one care planning meeting.39 Future pragmatic trials should consider such resource constraints when designing studies.
Throughout the implementation of PROVEN, the research team learned many lessons regarding the training of champions for ACP that should considered for future pragmatic trials focused on ACP in NHs. First, ACP champions must be trained in having ACP conversations. Second, senior leadership must have the expectation that ACP champions are indeed comfortable to have these conversations. Finally, many champions did not think that serious illness conversations were within their scope of responsibility. Notating whether or not a patient has an advance directive and completing a checklist was universally acknowledged as a task suited for their skillsets. However, having a goals-of-care conversation or other ACP conversation was beyond many of the champions’ comfort level.
The NH population is heterogeneous, making care difficult and complex.40 In our supplementary analysis of short-stay patients without advanced illness stratified by mortality risk, we found large differences in hospitalization rates. Over half of the patients in the highest mortality risk group were transferred to hospital at least once in the 100-day follow-up period, which was 10 percentage points higher than patients at low risk of dying. Given the high mortality coupled with the high hospitalization risk in these groups, it should not be surprising that hospice enrollment was low; 23% among those with advanced illness and 13% among those without advanced illness in the high MRS group. This low uptake of hospice is consistent with previous reports of trends among advanced illness NH residents.41 Although hospice may be more appropriate for short-stay patients near end of life, patients cannot receive care under both the Medicare skilled nursing facility benefit and the Medicare hospice benefit. As NHs are reimbursed higher under the skilled nursing facility benefit, NHs are disincentivized from referring patients to hospice, when appropriate.42,43
Embedding large-scale intervention studies into the real world means that external events such as policy changes can have an independent influence on study outcomes.44 During the study period, there were several concurrent national policies focused on reducing hospitalizations from NHs. In 2016, Nursing Home Compare started reporting three new quality measures focused on short-stay patients: rehospitalizations, emergency department visits, and successful discharge to the community. PROVEN’s outcome include two of these measures in its hospital transfer rate. Initial reports examining these measures found that rehospitalization rates declined by 6.7%.4 Although these secular trends are unlikely to differentially impact the trial arms, reducing hospital transfers over and above the national declines could be much more difficult.
The findings of this study should be considered in the context of several limitations. The primary outcome cohort for the PROVEN trial was not the short-stay population, and efforts to improve fidelity were focused on the long-stay population. The follow-up time for this study population was only 100 days, it may be that a person could be exposed to the ACP video at a healthy stage where they start thinking about it and then the next few years or so after a health status change their preferences toward being hospitalized may change. Finally, although the ACP videos may stimulate conversation about choices at end of life,45 most NHs had low show-rates and varied substantially. NHs found it difficult to incorporate and sustain as part of standard operating procedures,37 so it is difficult to determine whether the ACP video would be effective if implementation was improved.
CONCLUSIONS
An ACP video program did not significantly reduce hospital transfers, burdensome treatment, or hospice enrollment, among short-stay patients. Given the constrained resources and competing priorities faced by NHs, a more practical approach to offering additional tools for ACP could involve targeting those at highest risk of hospitalization instead of everyone at admission.
ACKNOWLEDGMENTS
This work is supported within the National Institutes of Health (NIH) Health Care Systems Research Collaboratory by the NIH Common Fund through cooperative agreement U24AT009676 from the Office of Strategic Coordination within the Office of the NIH Director and cooperative agreement UH3AG049619 from the National Institute on Aging (NIA). Dr Mitchell is supported by the NIA under award number K24AG033640. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Conflict of Interest: Dr Mor chairs the Scientific Advisory Committee for naviHealth, a postacute care convener serving hospitals and Medicare advantage plans, a role for which he is compensated. Dr Loomer was previously a consultant for the American Health Care Association. Dr Gutman reports consulting for Johnson & Johnson/Janssen.
Sponsor’s Role: The funding sources for this study played no role in the design or conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.
REFERENCES
- 1.Medicare Payment Advisory Commission. Skilled nursing facility services. 2019. http://www.medpac.gov/docs/default-source/reports/mar19_medpac_ch8_sec.pdf
- 2.Wang S-Y, Aldridge MD, Gross CP, et al. Transitions between healthcare settings of hospice enrollees at the end of life. J Am Geriatr Soc 2016;64(2): 314–322. 10.1111/jgs.13939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Flint LA, David DJ, Smith AK. Rehabbed to death. N Engl J Med 2019;380(5):408–409. 10.1056/NEJMp1809354. [DOI] [PubMed] [Google Scholar]
- 4.Saliba D, Weimer DL, Shi Y, Mukamel DB. Examination of the new short-stay nursing home quality measures: rehospitalizations, emergency department visits, and successful returns to the community. Inquiry. 2018;55:1–11. 10.1177/0046958018786816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Temkin-Greener H, Zheng NT, Xing J, Mukamel DB. Site of death among nursing home residents in the United States: changing patterns, 2003–2007. J Am Med Dir Assoc 2013;14(10):741–748. 10.1016/j.jamda.2013.03.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Allers K, Hoffmann F, Schnakenberg R. Hospitalizations of nursing home residents at the end of life: a systematic review. Palliat Med 2019;33(10): 1282–1298. 10.1177/0269216319866648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gozalo PL, Miller SC. Hospice enrollment and evaluation of its causal effect on hospitalization of dying nursing home patients. Health Serv Res 2007;42 (2):587–610. 10.1111/j.1475-6773.2006.00623.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Miller SC, Intrator O, Gozalo P, Roy J, Barber J, Mor V. Government expenditures at the end of life for short- and long-stay nursing home residents: differences by hospice enrollment status. J Am Geriatr Soc 2004; 52(8):1284–1292. 10.1111/j.1532-5415.2004.52357.x. [DOI] [PubMed] [Google Scholar]
- 9.Ouslander JG, Naharci I, Engstrom G, et al. Root cause analyses of transfers of skilled nursing facility patients to acute hospitals: lessons learned for reducing unnecessary hospitalizations. J Am Med Dir Assoc 2016;17(3): 256–262. 10.1016/j.jamda.2015.11.018. [DOI] [PubMed] [Google Scholar]
- 10.Ouslander JG, Naharci I, Engstrom G, et al. Lessons learned from root cause analyses of transfers of skilled nursing facility (SNF) patients to acute hospitals: transfers rated as preventable versus nonpreventable by SNF staff. J Am Med Dir Assoc 2016;17(7):596–601. 10.1016/j.jamda.2016.02.014. [DOI] [PubMed] [Google Scholar]
- 11.Jennings LA, Ramirez KD, Hays RD, Wenger NS, Reuben DB. Personalized goal attainment in dementia care: measuring what persons with dementia and their caregivers want. J Am Geriatr Soc 2018;66(11):2120–2127. 10.1111/jgs.15541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Palan Lopez R, Mitchell SL, Givens JL. Preventing burdensome transitions of nursing home residents with advanced dementia: it’s more than advance directives. J Palliat Med 2017;20(11):1205–1209. 10.1089/jpm.2017.0050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mitchell SL, Teno JM, Intrator O, Feng Z, Mor V. Decisions to forgo hospitalization in advanced dementia: a nationwide study. J Am Geriatr Soc 2007;55(3):432–438. 10.1111/j.1532-5415.2007.01086.x. [DOI] [PubMed] [Google Scholar]
- 14.Volandes AE, Paasche-Orlow MK, Barry MJ, et al. Video decision support tool for advance care planning in dementia: randomised controlled trial. BMJ 2009;b2159:338. 10.1136/bmj.b2159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Volandes AE, Paasche-Orlow MK, Mitchell SL, et al. Randomized controlled trial of a video decision support tool for cardiopulmonary resuscitation decision making in advanced cancer. J Clin Oncol 2013;31(3):380–386. 10.1200/JCO.2012.43.9570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.El-Jawahri A, Podgurski LM, Eichler AF, et al. Use of video to facilitate end-of-life discussions with patients with cancer: a randomized controlled trial. J Clin Oncol 2010;28(2):305–310. 10.1200/JCO.2009.24.7502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Volandes AE, Ferguson LA, Davis AD, et al. Assessing end-of-life preferences for advanced dementia in rural patients using an educational video: a randomized controlled trial. J Palliat Med 2011;14(2):169–177. 10.1089/jpm.2010.0299. [DOI] [PubMed] [Google Scholar]
- 18.Volandes AE, Brandeis GH, Davis AD, et al. A randomized controlled trial of a goals-of-care video for elderly patients admitted to skilled nursing facilities. J Palliat Med 2012;15(7):805–811. 10.1089/jpm.2011.0505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Volandes AE, Lehmann LS, Cook EF, Shaykevich S, Abbo ED, Gillick MR. Using video images of dementia in advance care planning. Arch Intern Med 2007;167(8):828–833. 10.1001/archinte.167.8.828. [DOI] [PubMed] [Google Scholar]
- 20.Volandes AE, Barry MJ, Chang Y, Paasche-Orlow MK. Improving decision making at the end of life with video images. Med Decis Making. 2010;30(1): 29–34. 10.1177/0272989X09341587. [DOI] [PubMed] [Google Scholar]
- 21.Mitchell SL, Shaffer ML, Cohen S, Hanson LC, Habtemariam D, Volandes AE. An advance care planning video decision support tool for nursing home residents with advanced dementia: a cluster randomized clinical trial. JAMA Intern Med 2018;178(7):961–969. 10.1001/jamainternmed.2018.1506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Mitchell SL, Volandes AE, Gutman R, et al. Advance care planning video intervention among long-stay nursing home residents: a pragmatic cluster randomized clinical trial. JAMA Intern Med 2020;180:1070–1078. 10.1001/jamainternmed.2020.2366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mor V, Volandes AE, Gutman R, Gatsonis C, Mitchell SL. PRagmatic trial Of Video Education in Nursing homes: the design and rationale for a pragmatic cluster randomized trial in the nursing home setting. Clin Trials 2017; 14(2):140–151. 10.1177/1740774516685298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Centers for Medicare and Medicaid Services. Long-Term Care Facility Resident Assessment Instrument 3.0 User’s Manual. Baltimore, MD: Department of Health and Human Services; 2019. https://downloads.cms.gov/files/mds-3.0-rai-manual-v1.17.1_october_2019.pdf. Accessed June 27, 2020. [Google Scholar]
- 25.Thomas KS, Dosa D, Wysocki A, Mor V. The minimum data set 3.0 Cognitive Function Scale. Med Care. 2017;55(9):e68–e72. 10.1097/MLR.0000000000000334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Morris JN, Fries BE, Morris SA. Scaling ADLs within the MDS. J Gerontol A Biol Sci Med Sci 1999;54(11):M546–M553. 10.1093/gerona/54.11.m546. [DOI] [PubMed] [Google Scholar]
- 27.Thomas KS, Ogarek JA, Teno JM, Gozalo PL, Development MV. Validation of the nursing home minimum data set 3.0 mortality risk score (MRS3). J Gerontol A Biol Sci Med Sci 2019;74(2):219–225. 10.1093/gerona/gly044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mitchell SL, Teno JM, Kiely DK, et al. The clinical course of advanced dementia. N Engl J Med 2009;361(16):1529–1538. 10.1056/NEJMoa0902234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gozalo P, Teno JM, Mitchell SL, et al. End-of-life transitions among nursing home residents with cognitive issues. N Engl J Med 2011;365(13):1212–1221. 10.1056/NEJMsa1100347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Mitchell SL, Teno JM, Roy J, Kabumoto G, Mor V. Clinical and organizational factors associated with feeding tube use among nursing home residents with advanced cognitive impairment. JAMA. 2003;290(1):73–80. 10.1001/jama.290.1.73. [DOI] [PubMed] [Google Scholar]
- 31.Givens JL, Jones RN, Shaffer ML, Kiely DK, Mitchell SL. Survival and comfort after treatment of pneumonia in advanced dementia. Arch Intern Med 2010;170(13):1102–1107. 10.1001/archinternmed.2010.181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Miller SC, Lima JC, Looze J, Mitchell SL. Dying in U.S. nursing homes with advanced dementia: how does health care use differ for residents with, versus without, end-of-life Medicare skilled nursing facility care? J Palliat Med 2012;15(1):43–50. 10.1089/jpm.2011.0210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Teno JM, Gozalo P, Khandelwal N, et al. Association of increasing use of mechanical ventilation among nursing home residents with advanced dementia and intensive care unit beds. JAMA Intern Med 2016;176(12):1809–1816. 10.1001/jamainternmed.2016.5964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Lee AH, Wang K, Scott JA, Yau KK, McLachlan GJ. Multi-level zero-inflated Poisson regression modelling of correlated count data with excess zeros. Stat Methods Med Res 2006;15(1):47–61. [DOI] [PubMed] [Google Scholar]
- 35.Davison AC, Hinkley DV. Bootstrap Methods and Their Application. Cambridge: Cambridge University Press; 1997. [Google Scholar]
- 36.Loomer L, McCreedy E, Belanger E, et al. Nursing home characteristics associated with implementation of an advance care planning video intervention. J Am Med Dir Assoc 2019;20(7):804-809.e1. 10.1016/j.jamda.2019.01.133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Palmer JA, Parker VA, Barre LR, et al. Understanding implementation fidelity in a pragmatic randomized clinical trial in the nursing home setting: a mixed-methods examination. Trials. 2019;20(1):656. 10.1186/s13063-019-3725-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Palmer JA, Parker VA, Mor V, et al. Barriers and facilitators to implementing a pragmatic trial to improve advance care planning in the nursing home setting. BMC Health Serv Res 2019;19(1):527. 10.1186/s12913-019-4309-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Quinn CC, Port CL, Zimmerman S, et al. Short-stay nursing home rehabilitation patients: transitional care problems pose research challenges. J Am Geriatr Soc 2008;56(10):1940–1945. 10.1111/j.1532-5415.2008.01852.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ouslander JG, Grabowski DC. Rehabbed to death reframed: in response to “Rehabbed to death: breaking the cycle”. J Am Geriatr Soc 2019;67(11): 2225–2228. 10.1111/jgs.16127. [DOI] [PubMed] [Google Scholar]
- 41.McCarthy EP, Ogarek JA, Loomer L, et al. Hospital transfer rates among US nursing home residents with advanced illness before and after initiatives to reduce hospitalizations. JAMA Intern Med 2019;180:385–394. 10.1001/jamainternmed.2019.6130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Huskamp HA, Stevenson DG, Chernew ME, Newhouse JP. A new Medicare end-of-life benefit for nursing home residents. Health Aff 2010;29(1):130–135. 10.1377/hlthaff.2009.0523. [DOI] [PubMed] [Google Scholar]
- 43.Stevenson DG, Bramson JS. Hospice care in the nursing home setting: a review of the literature. J Pain Symptom Manage. 2009;38(3):440–451. 10.1016/j.jpainsymman.2009.05.006. [DOI] [PubMed] [Google Scholar]
- 44.Curtis LH, Dember LM, Vazquez MA, et al. Addressing guideline and policy changes during pragmatic clinical trials. Clin Trials. 2019;16(4):431–437. 10.1177/1740774519845682. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Loomer L, McCreedy E, Belanger E, et al. Proxies viewing decision support video in nursing home report higher advance care planning engagement. J Am Med Dir Assoc 2019;20(11):1467–1469. 10.1016/j.jamda.2019.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]