Results
We observed staff activity in four wards in both the POD programme implementation phase and delivery phase. We observed one ward in the implementation phase only, as it underwent a change of specialty before the delivery phase, meaning that observations in the delivery period would not have made a valid comparison possible.
We undertook 8257 10-minute observations in the implementation phase and 6711 10-minute observations during the delivery phase: 14,968 observation in total. This equates to almost 2500 hours of staff time.
Data validity and reliability
The amount of nursing attention given to patients should be directly proportional to their dependency, with more dependent patients, on average, receiving more care (‘care ratios’). In our study, the highest-dependency patients (grade 4) received 13 times (grade 4/grade 1 = 9/0.7) and 14 times (grade 4/grade 1 = 8.4/0.6) more nursing care in the pre-POD and post-POD periods, respectively (). Any deviation from these incremental rising care times may indicate that patients’ dependency levels have been misattributed or that the non-participant observers have labelled interventions inaccurately. Either of these would be a threat to the validity and reliability of the data. Generally, the ‘care ratios’ for the wards were sound in this study, showing accuracy and consistency of data in both the pre-POD and post-POD implementation periods.
Direct care ratios (minutes per hour)
Average ward occupancy on all wards was similar in both periods of data collection (implementation phase, 27.9 occupied beds; delivery phase, 28.2 occupied beds) ().
Ward occupancy, patient dependency and workload
The majority of patients in both phases of the study were either dependent or highly dependent (pre POD delivery: 71% dependent or highly dependent; post POD delivery: 73% dependent or highly dependent).
The workload index for the wards was similar in the pre-POD and post-POD delivery periods (workload indices 3.2 and 3.3, respectively).
The percentages of staff at different grades observed on duty during the data collection periods were similar at both time points (see Table 6).
Ward sister or manager grade staff accounted for 6% of the observations at both time points; staff nurses accounted for 45% and 46% of the observations pre POD and post POD, respectively. Support workers accounted for almost half of the observations (49% pre POD implementation, 46% post POD implementation). Volunteers accounted for only 1% of the post-POD delivery observations (see Table 6).
There were small changes overall in direct and indirect patient care from the pre-POD to the post-POD observations (direct patient care: 45% pre POD to 46% post POD; indirect patient care: 28% pre POD to 29% post POD). There was a 4% increase in direct patient care by ward sister and staff nurse grades, whereas there was a 2% decrease in direct patient care in support worker grades (see Table 6). Overall, there were also small decreases between the implementation and delivery phases in the precentage of both associated work (from 15% to 13%) and personal time (from 13% to 12%).
Discussion
There was a modest increase in the percentage of time spent by staff in both direct and indirect care following the introduction of the POD intervention. An increase in both direct care and indirect care is unusual, as they usually demonstrate an inverse relationship; that is, if one activity increases in a ward, then the other usually falls. The reason for this is not clear.
The data support our hope that the implementation of POD would not be associated with adverse effects on nurse workload, which could have been a consequence of introducting a system of enhanced care such as the POD programme. Indeed, there was an indication that the introduction of the POD programme on the wards was associated with a small positive change overall from associated care and personal time to direct and indirect care.
An assessment of these changes to the staff workload data of the wards before and after the introduction of the POD programme to the wards is potentially vulnerable to a number of factors, including the validity and reliability of the data and the changes to bed occupancy and patient dependency. Staff activity and workload in wards are partly driven by the bed occupancy rates and patient dependency data. The data show that the pre- and post-POD data for bed occupancy and patient dependeny were similar. Therefore, comparing staff activity in pre- and post-POD periods was meaningful, that is the ward workload was not a major confounding variable. The validity and reliability of the data were also acceptable: there was consistency across the two time points. Any changes in the workload data can, therefore, reasonably be attributed to the implementation of the POD system of care.
