Educational Video to Improve Nursing Home Care in End-stage Dementia (EVINCE)
Behavioral: Advance care planning intervention
Other: Control group
|Study Design:||Allocation: Randomized
Endpoint Classification: Efficacy Study
Intervention Model: Parallel Assignment
Masking: Double Blind (Investigator, Outcomes Assessor)
|Official Title:||Educational Video to Improve Nursing Home Care in End-stage Dementia|
- Documented decisions to forgo hospitalization [ Time Frame: by six months ] [ Designated as safety issue: No ]
The proportion of residents with this outcome will be considered cumulatively at 6 months, including those who died; i.e., a composite of the percent of residents alive at six months who had a decision not to hospitalize and those who died before six months with this outcome prior to death. This cumulative outcome will only be based on time points following baseline as the baseline chart review data is conducted before the baseline proxy interview (i.e., before the proxy has seen the video in the intervention arm or heard options verbally in the control arm).
(Decisions to forehospitalizations will be examined in a similar fashion at 3, 9, and 12 months, however the six month time frame is the primary trial outcome)
- Acquisition of decisions not to hospitalize [ Time Frame: by 12 months ] [ Designated as safety issue: No ]This is a sub primary outcome. The modified ITT population will be the subgroup of residents who begin the study without a documented decision to forego hospitalization, and the outcome will be acquisition of a documented decision to forego hospitalization over the 12 month follow-up period. The analysis will utilize Cox proportional hazards regression. Results will be summarized using a hazard ratio and associated 95% confidence interval as well as plots of the cumulative incidence by group.
- Preference for level of care [ Time Frame: baseline, 3, 6, 9, and 12 months ] [ Designated as safety issue: No ]The proportion of proxies who have chosen comfort care (versus intermediate or intensive care) based on telephone interviews will be considered cumulatively at each assessment. For example, at 6 months the outcome will include the proportion of proxies choosing comfort care up to and including the 6 month interview. Cumulative proportions will include data from the baseline interviewfor the control group, and baseline immediate post video interview for intervention group.The proportion of proxies choosing comfort care will be compared between the intervention and control groups using an extension of logistic regression based on general estimating equations (GEE) to account for clustering at the NH level at each time period.
- Acquistion of a preference for comfort care. [ Time Frame: by 12 months ] [ Designated as safety issue: No ]This is a sub-analysis for outcome 3. It will examine the acquisition of a preference for comfort care by 12 months in the subset of residents for whom the proxy had not chosen comfort care at baseline. Only residents who survive to the 3 month proxy interview will be included as no proxy post-death interview is conducted and therefore data are not available describing preferences beyond baseline.
- Other documented advance care planning [ Time Frame: 3, 6, 9, 12 months ] [ Designated as safety issue: No ]Comparison of documentation of advance care planning at each assessment period between residents in the intervention and control arms, including: i. proxy decisions to forgo burdensome interventions ( tube-feeding, antimicrobial use, and parenteral therapy) and ii. goals of care discussions between proxies and providers. Decisions to withhold each of the aforementioned interventions will be examined as separate outcomes. The analytic approach will be the same as that described for Outcome 1.
- Acquisition of other documented advance care planning [ Time Frame: by 12 months ] [ Designated as safety issue: No ]Analyses will include the subgroup of residents who begin the study without previously documented: i. decisions to forego each of the specified burdensome treatments, and ii. goals of care discussions. The analytic approach will be the same as described for Outcome 2.
- Hospital transfers and other burdensome interventions [ Time Frame: by 12 months ] [ Designated as safety issue: No ]Burdensome treatments include: i. Hospitals transfers (hospitalizations or emergency room visits) and ii. tube feeding or parenteral therapy. Analyses will focus on two prespecified outcomes: i. the rate of hospitalization or ED visit over the entire 12 month follow-up period and ii. the rate of residents experiencing any hospitalization, ED visit, or parenteral therapy over the entire 12 month follow-up period. These outcomes will be compared between the intervention and control arms.
- Hospice enrollment [ Time Frame: By 12 months ] [ Designated as safety issue: No ]Residents on hospice at baseline will be excluded from these analyses. The outcome will be the cumulative incidence of hospice enrollment over the 12 month follow-up period and will be compared among residents in the intervention and control arms.
|Study Start Date:||March 2013|
|Estimated Study Completion Date:||February 2018|
|Estimated Primary Completion Date:||September 2017 (Final data collection date for primary outcome measure)|
Experimental: Advance care planning intervention
At baseline, health care proxies in the intervention arm will be shown a 12-minute Advance Care planning video that describes 3 levels of treatment in advanced dementia: comfort basic and intensive. After viewing the video, the proxies will be asked their preferred level of care for the resident and this choice will be communicated to the residents primary care team in a written form.
|Behavioral: Advance care planning intervention|
Active Comparator: Usual care
Residents in control nursing homes with receive the usual advance care planning that occurs in their nursing home.
|Other: Control group|
The over-riding goal of the EVINCE study is to conduct a cluster RCT of a video Advance Care Planning intervention in nursing home residents with advanced dementia. A total of 360 nursing home residents with advanced dementia (N=180/arm) in 20 matched nursing homes (10 intervention/10 control) will be recruited. These residents' proxies with also be recruited for a total sample size of 720 residents and proxies (resident/proxy dyads). Outcomes will be compared between residents in the intervention vs, control nursing homes at baseline and every 3 months up to 12 months. At baseline proxies of residents in the intervention nursing homes will view a 12-minutes video describing three level of care options in advanced dementia (intensive, basic and comfort) and their choice will be communicated to the resident's primary care providers. Residents in the control nursing homes will receive the usual advance care planning practiced in those facilities.
UPDATE SEPT 19, 2014: During the first 18 months of the study, recruitment of resident/proxy dyads was lower per facility than anticipated at the time of study design. Therefore, the number of matched pairs of NH recruited was increased. As of September 19, 2014, 19 NH pairs (38 NHs) were recruited and randomized with a plans to continue to up to 30 pairs as needed.
NHs were originally randomized using a paired approach matched for for-profit status and whether or not the NH had a special care dementia unit (SCU). In March 2014, the state of Massachusetts changed legislation defining an SCU. Many of the changes focused on specifics of staff training. As a result, SCUs in several participating facilities did not meet the new criteria and lost this official designation, although the actual clinical structure of the existing units did not change. Nonetheless, we opted to maintain the initial matching criteria of the first 19 NHs which included SCU based on the definition before the Massachusetts legislation change, but dropped this matching criterion after the legislation change. Thus, beginning with matched pair 20, NHs were matched solely on for-profit status.
UPDATE MARCH 2016 To reach subject recruitment numbers, additional facilities were recruited. As of this date, 60 NHs (30/arm) have been recruited, randomized and have had subject enrolled.
To achieve additional for power for Outcome 2, (aquisition of new decisions to forego hospitalizations) the target sample size was increased to 400.
Please refer to this study by its ClinicalTrials.gov identifier: NCT01774799
|United States, Massachusetts|
|Hebrew Rehabilitation Center|
|Boston, Massachusetts, United States, 02131|
|Principal Investigator:||Susan L Mitchell, MD, MPH||Hebrew Rehabilitation Center, Boston|
|Principal Investigator:||Angelo Volandes, MD, MPH||Massachusetts General Hospital|