Response Bias and Generalizability
Because the survey was open to all eligible NYS nursing homes rather than based on a randomly selected sample of facilities, the possibility of a response bias must be considered. To assess this, we compared the 149 nursing homes included in the analytical sample with all eligible NYS facilities on several characteristics that could affect presence of daily practice teams (). The participating homes appear to be statistically significantly different (p = .03) in terms of occupancy rates (94.2% vs. 93.1%) from all NYS facilities eligible for the study, but this difference does not seem to be operationally or organizationally meaningful. However, the participating homes are significantly (p < .0001) less likely to represent for-profit facilities in New York (31.5% vs. 49.1%). We correct for this potential bias in the multivariate analysis with the use of sampling probability weights. We find no other statistically significant (p < .05) differences between the study sample and all eligible nursing homes on the remaining facility characteristics or with regard to quality of care and staffing measures.
Characteristics of Sample Facilities Compared With All Eligible NYS Nursing Homes
Description of Daily Practice Teams
Descriptive statistics consisting of the percent and sample size (categorical variables) and the mean, standard deviation, and range (continuous variables) are presented for the dependent and the independent variables in .
Variables Included in the Analysis of Daily Care Teams Prevalence: Descriptive Statisticsa
In an average facility, 15.9% of the direct care workers view themselves as working in formally organized daily practice teams. The distribution of daily practice team penetration in the study nursing homes is shown in . In half of the facilities, daily practice team penetration is lower than 15%, but only a few (n = 3) report no daily practice teams at all. Similarly, almost half of the facilities report team prevalence greater than 20%, but none report penetrations above 44%. It is important to point out that these daily practice teams are reported to be formally organized and to be multidisciplinary (). Overall, the CNAs report 84% of their daily practice teams to include other disciplines, whereas the LPNs and the RNs report, respectively, 94% and 98% of their teams to be multidisciplinary. These workers report roughly similar participation of their own professions in the daily practice teams but show more variation with regard to team participation of physicians, social workers, and therapists (). For example, whereas 46% of the CNAs report physicians to be on their teams, 66% and 69% of the LPNs and the RNs, respectively, report physicians on daily practice teams.
Prevalence of daily practice teams in nursing homes (N = 149).
Daily practice team composition reported by CNAs, RNs, and LPNs. CNA = certified nurse assistant; LPN = licensed practical nurse; RN = registered nurse; SW = social worker; PT/OT = therapist; MD = physician; Or = other.
The respondents were also asked about specific activities such as team meetings with other members to discuss residents’ care. Almost half (47.2%) report these meetings to occur daily, whereas an additional 25.0% and 16.2% report meeting at least once per week or once per month, respectively.
In at least 70% of the facilities, NHAs (71.8%) and DONs (77.2) report presence of formal daily practice teams. The NHAs and the DONs are very enthusiastic about the importance of teamwork in the operations of their facilities. Almost all view teams as very important in promoting residents' quality of life (89.6%) and clinical quality of care (90.3%). Furthermore, 79.2% report teams as very important in maintaining good staff–resident relationships, and 68.2% think teams are very important with regard to staff turnover and retention ().
Description of other independent variables is also provided in . Nursing home staff assessed close to 80% of the facilities as having collegial (47.65%) or supportive (30.87%) management styles, that is, not hierarchical and open to employee input, whereas an additional 21.48% reported homes with more traditional and authoritarian management style (i.e., custodial/autocratic).
There appears to be substantial variability across nursing homes with regard to the job design indicators. In an average facility, 40% of the direct care workers report having primary assignment (SD = 15.9%), ranging from none to 100% across all facilities. In the average facility, 33.2% of direct care workers report participating on teams other than the daily practice teams (SD = 12%), with a range from 8.7% to 83.3%.
There is also substantial variability across these facilities in both tenure and turnover of the top administrators. For example, the average tenure is 81 months (SD = 73) for the DONs and 169 months (SD = 113) for the NHAs.
Characteristics of Nursing Homes With Daily Practice Teams
We estimated a linear regression model with robust standard errors and sampling weights, correcting for response bias, to examine the characteristics of nursing homes associated with daily practice teams' prevalence. (The model was also estimated without the probability weights, and the results, not shown here, were virtually unchanged.) The results are summarized in
Characteristics Predicting Prevalence of Daily Care Teams in Nursing Homes: Linear Regression Model With Robust Standard Errors and Probability Weightsa
The importance of several management practices in predicting prevalence of daily practice teams supports several posited hypotheses. In facilities where nursing home leaders consider teams very important with regard to clinical quality of care, team prevalence is significantly higher by 3.3% (Hypothesis 1b). However, administrators’ perceptions of the importance of teams with regard to other aspects of care (e.g., residents’ quality of life), reported as equally important, are not significantly associated with the presence of daily practice teams. Furthermore, team prevalence as reported by the direct care staff is 2.6% higher in facilities in which the DON reports presence of formal daily practice teams (Hypothesis 1c).
Of the two job design variables—primary assignment and participation on other teams—only the latter is statistically significant (p < .001). A 10% increase in direct care staff involvement in other than daily teams is associated with a 2.5% increase in prevalence of daily practice teams (Hypothesis 2a). This is the strongest predictor (standardized coefficient = 41.20) of daily practice team prevalence in the study facilities.
One of the facility characteristics is also shown to be significantly associated with daily practice team prevalence, providing support for Hypothesis 4a. A 1-hr increase in nursing hours per resident per day is associated with a 1.95% increase in daily team prevalence.
We find no statistically significant association between managers’ characteristics or management practices and team prevalence.