Observational study of the capacity and demand of plain-film workflow in a radiology department

ObjectivesDemand for radiology services in the National Health Service (NHS) is rising. The pressure felt by radiology departments is compounded by growing public expectation, government guidelines, targets, penalties and tight budget restrictions. One widely supported hypothesis is that inefficiency in the NHS is a result of a mismatch between the variances in capacity and demand. In the light of an increasing trend towards evidence-based management the study aims to model, analyse and understand variations in plain-film workflow in a radiology department and evaluate whether the data provide evidence to base future decisions upon.MethodsRetrospective data for a period of 6 months were collected, from the Computer Radiology Information System (CRIS), staff rotas and clinic diaries. Capacity was measured by the planned number of radiographers working within the department. Demand was measured by the daily workload of the department, number of plain-film events, and was subdivided to include referral source. To further analyse the drivers for demand the number of outpatient clinics was also recorded. Descriptive statistical testing was used to understand the variability in the obtained data. Levene's test was undertaken to test the homogeneity of daily variances in clinics and workload. Establishment of correlative relationships was undertaken using Pearson Product Moment Correlation (r) between chosen variables. Linear regression testing was performed in order to establish the capacity of the number of clinics running to predict the workload, adjusted for GP events, of the department.ResultsMean daily workload, capacity and clinics show variable correlation. Workload and clinics demonstrate relatively high variation; workload; range, max = 178, min = 46, mean = 95.58, standard deviation (SD) = 25.35, coefficient of variation (CV) = 0.27. Clinics; mean = 4.79, SD = 1.63, CV = 0.34. Variances in daily clinics and daily workflow are homogeneous, Levene tests F = 1.683, p = 0.159 and F = 0.67109, p = 0.614, respectively. There was a low degree of positive correlation between workload and capacity r = 0.322, p < 0.01. There was a moderate degree of negative correlation between clinics and GP examinations r = −0.486, p < 0.00. There was a marked degree of positive correlation between clinics and workload, when workload has been adjusted for GP events r = 0.686, p < 0.00. The linear regression suggested that 47% of the variance in GP-adjusted workload could be explained by the number of clinics (r2 = 0.471, F = 95.201, p < 0.00).ConclusionsThe results suggest potential hazards in the current intuitive approach to workflow management. The study suggests two practical approaches for management in applying the findings. The results support the concepts of evidence-based management and a global patient-pathway process mindset.