Pressure ulcer multi-disciplinary teams via telemedicine (PUMTT) : 2009 CPSI research competition : final report

Additional parameters of the simulation model included the measurement error of normalized wound surface areas (0.1 standard deviation units), the percentage of ulcers that were not likely to respond to the intervention (20%), the estimated prevalence of Stage 2-4 ulcers (4%), and the estimated annual incidence of Stage 2-4 pressure ulcers (2.5%). [...] Time to Healing From the 91 PUs identified during the control period, 28(30.8%) healed in the control period, and 3(3.32%) were censored at the end of the control period. [...] First, we observe the large proportion of censored observations (n=107, 53.2%), which reduces the effective sample size of the analysis, potentially introducing bias for the estimate of the time to healing, resulting in potentially inaccurate approximations of the variances used in the analysis. [...] Additionally, any occurrence of uneven censoring due to uneven lengths of observations time between the two periods may potentially introduce bias in the estimation of the difference of the healing time between the two periods. [...] To investigate the effect of the intervention we fitted a random effects model for the estimation of the incidence rate ratio (IRR) between control and intervention.