Background
Importance of residential long term care
Ongoing challenges to quality of care and use of best practices
The Translating Research in Elder Care (TREC) research program
Relevance of clinical microsystems
Relevance of the care unit work context
Purpose and aims
Methods/Design
Design
T-2 (09/14-04/15) | T-1 (10/15-11/15) | T0 (11/15-05/16) | T1 (05/16-06/16) | T1 (10/16-11/16) | T1 (04/17-05/17) | Tx (05/17-12/17) | Tx (01/19-06/19) | |
---|---|---|---|---|---|---|---|---|
Baseline | Dissemination workshop | Study preparation | Goal setting workshop | Support workshop 1 | Support workshop 2 | Post intervention | Long-term follow-up | |
Assessment of primary and secondary study outcomes (all facilities participating in TREC)
| ||||||||
TREC survey | X | X | X | |||||
RAI-MDS 2.0 | X | X | X | |||||
Unit profile survey | X | X | X | |||||
Facility profile survey | X | X | X | |||||
Enrolment (all TREC facilities in Alberta and British Columbia)
| ||||||||
Ethics/operational approvals | X | |||||||
Eligibility screening | X | |||||||
Randomization/allocation | X | |||||||
Recruitment and informed consent (BAF/EAF)
| X | |||||||
Study intervention | ||||||||
Standard Feedback (SF)
| X | |||||||
Basic Assisted Feedback (BAF)
| X | X | X (virtual) | X (virtual) | ||||
Enhanced Assisted Feedback (EAF)
| X | X | X (face-to-face) | X (face-to-face) | ||||
Ongoing phone/e-mail support (EAF)
| ––––––– | ––––––– | ––––––– | |||||
Process evaluation | ||||||||
Workshop evaluation surveys (at the end of each workshop) | X | X | X | X | ||||
Workshop fidelity checklist and debriefs (at the end of each workshop) | X | X | X | X | ||||
Detailed documentation of workshops (participant observations) | X (during workshops) | |||||||
Focus groups with BAF/EAF unit managers | X (05/2017) | X (12/2018) | ||||||
Semi-structured interviews with SF unit managers | X (01/2017) | |||||||
Basic cost evaluation of the intervention | ––––––– | ––––––– | ––––––– |
Intervention
Dissemination workshops
Goal Setting Workshops
Support workshops
On-demand e-mail and phone support
Sample
Sample size calculation
-
Yijt is the ACT Formal Interactions (FI) score (primary outcome) of unit j in facility i at time t.
-
μ is the FI population mean.
-
A1ijt, A2ijt, and A3ijt are indicator variables for the interventions (SF, BAF, and EAF, respectively). The indicator variable is 1 if the unit has been exposed to the respective intervention and 0 otherwise.
-
β1, β2, and β3 are the treatment effects of the three interventions.
-
bi is a facility-level random effect (variability of units within the facility).
-
bij is a unit-level random effect (variability of time points within the unit).
-
εijt is a residual term (variability between units).
-
Standard deviation of bi = 0.154 (ICC = 0.445) (FI variability of units, within facilities)
-
Standard deviation of bij = 0.104 (FI variability of units, across waves 1 and 2 in TREC 1.0)
-
Standard deviation of εijt = 0.192 (variability of the unit FI residual term)
-
Average cluster size = 3 units per facility
Inclusion and exclusion criteria
Inclusion criteria | Exclusion criteria | |
---|---|---|
Facilities | • Participation in the TREC observational study • Located in one of four health regions in Alberta (Edmonton or Calgary) or British Columbia (Fraser Health or Interior Health) • Have at least one care unit with ten or more care aide responses to our survey that we use to assess organizational context and staff outcomes at baseline (T-2) | • Does not participate in the TREC observational study • Located in the Winnipeg Regional Health Authority • Do not have at least one care unit with ten or more care aide responses to our survey that we use to assess organizational context and staff outcomes • Surveys collected in this facility cannot be assigned to a clinical microsystem as defined by TREC within this facility |
Care units | • Located within a facility participating in the TREC observational study in Alberta or British Columbia • Ten or more care aide responses to our survey that we use to assess organizational context and staff outcomes are available • A care manager can be identified who leads this unit | • Not located within a facility participating in the TREC observational study in Alberta or British Columbia • Less than ten care aide responses to our survey that we use to assess organizational context and staff outcomes are available • No care manager can be identified who leads this unit |