Skip to main content

Table 3 COVID-19 outbreaks by types of care homes and the impact of the tracker in COVID-19 containment

From: The use and impact of digital COVID-19 tracking in adult social care: a prospective cohort study of care homes in Greater Manchester

 

Outbreaks by home types

Tracker impact

Other drivers

 

Odd ratio

Odd ratio

Relative risk

Relative risk

Relative risk

Not-for-profit homes (vs for-profit homes)

2.46 [1.39,4.37]***

1.49 [0.90,2.47]

   

Residential homes (vs nursing homes)

0.38 [0.25,0.59]***

0.83 [0.56,1.22]

   

Small homes [1–23 occupied beds]

1.00 [=reference]

1.00 [=reference]

   

Medium homes [24–40 occupied beds]

0.76 [0.51,1.14]

0.82 [0.54,1.25]

   

Large homes [41+ occupied beds]

0.48 [0.27,0.86]**

0.54 [0.30,0.98]**

   

Chains owned homes (vs single-ownership homes)

2.09 [1.44,3.05]***

1.44 [1.02,2.04]**

   

Homes with residents with dementia (yes/no flag)

1.89 [1.26,2.83]***

1.21 [0.85,1.74]

   

Homes with learning disability and/or autism (yes/no flag)

0.19 [0.10,0.38]***

0.18 [0.09,0.34]***

   

Homes with residents with physical disabilities (yes/no flag)

1.08 [0.74,1.60]

1.17 [0.82,1.67]

   

Inadequate CQC rated

1.00

1.00

   

Requires improvement

6.00 [0.86,42.04]*

6.14 [1.13,33.42]**

   

Good

3.63 [0.52,25.19]

4.42 [0.82,23.95]*

   

Outstanding

4.01 [0.47,34.04]

2.71 [0.40,18.29]

   

Tracker uptake lag 7 (yes/no flag)

1.31 [0.81,2.13]

1.39 [0.87,2.22]

0.72 [0.42,1.25]

0.60 [0.35,1.04]*

0.72 [0.45,1.13]

Number of MSOA level COVID-19 cases lag 7

 

1.01 [1.00,1.02]***

 

1.06 [1.04,1.08]***

1.05 [1.03,1.07]***

Number of workers available lag 7

 

0.99 [0.98,1.00]

 

0.98 [0.97,0.99]***

0.99 [0.98,1.00]

Accepting new admissions [no issue] lag 7

 

1.00

 

1.00

1.00

Accepting new admissions [limited capacity] lag 7

 

1.38 [1.05,1.82]**

 

1.25 [0.85,1.83]

1.21 [0.83,1.76]

Accepting new admissions [emergencies only] lag 7

 

2.61 [1.21,5.62]**

 

0.86 [0.31,2.41]

0.90 [0.44,1.83]

Accepting new admissions [not possible] lag 7

 

2.01 [1.56,2.58]***

 

1.72 [1.19,2.50]***

1.56 [1.10,2.21]**

Worker RAG [Green] lag 7

 

1.00

 

1.00

1.00

Worker RAG [Amber] lag 7

 

1.43 [1.01,2.03]**

 

2.11 [1.68,2.66]***

1.18 [0.92,1.51]

Worker RAG [Red] lag 7

 

2.58 [1.39,4.79]***

 

4.28 [3.07,5.97]***

1.41 [0.99,2.01]*

Number of staff self-isolating lag 7

 

1.18 [1.11,1.25]***

  

1.29 [1.22,1.36]***

PPE [less than a week] lag 7

 

1.00

  

1 .00

PPE [1 to 4 weeks] lag 7

 

0.71 [0.49,1.02]*

  

0.74 [0.55,0.99]**

PPE [more than 1 month] lag 7

 

0.82 [0.49,1.39]

  

0.88 [0.48,1.61]

Number of beds available lag 7

   

1.01 [1.00,1.02]

1.00 [0.99,1.02]

Number of occupied beds lag 7

   

0.98 [0.97,1.00]**

0.99 [0.97,1.00]**

Home random effects

\(\checkmark\)

\(\checkmark\)

   

Home fixed effects

  

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

LA fixed effects

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

Week fixed effects

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

N

161,999

159,492

95,836

95,033

95,026

Homes

497

497

   

Chi-Squared

665.1 (p-val=0.00)

923.2 (p-val=0.00)

   

Pseudo R-Squared

  

0.318

0.380

0.433

Homes adopting tracker (estimation sample)

  

83

83

83

Homes not adopting tracker (estimation sample)

  

327

327

327

  1. 95% clustered robust confidence intervals are in the brackets. The dependent variable in Columns 1 and 2 is COVID-19 outbreaks [=1 if at least one resident has COVID-19 symptoms or tested positive on a given day, 0 otherwise]. The dependent variable in Columns 3–5 is COVID-19 symptomatic/positive residents per 100 beds. Column 1 estimated with random effect logistic, Column 2 with correlated random effects logistic models and Columns 3–5 with fixed effects Poisson regression (coefficients are incidence rate ratios). The estimations in Columns 1 and 2 also include a categorical variable on home local area index of multiple deprivation; the coefficients are mostly insignificant and omitted to save space. Column 4 does not control for any confounding factors. Column 5 includes factors other than PPE and staff self-isolating as these might be caused by the tracker uptake. All the control variables are included with lag 7 keeping in view the incubation period of COVID-19. Details on methods selection, missing data, and further sensitivity analysis are given in the Additional file 1: Appendices S3 and S4. *** p<0.01, ** p<0.05, * p<0.1