Respiratory Care Work Assignment Based on Work Rate Instead of Work Load

Discussion

A recent literature review found that decreased nurse staffing is associated with adverse outcomes in intensive care unit patients.⁴ There are also data to suggest that there is an association between staffing and teamwork. The higher the perception of the adequacy of staffing and the fewer patients cared for on a previous shift, the higher the teamwork scores.⁵ However, little is available in the literature regarding how staffing matches work load variability in healthcare. One study of nursing practice used process-control theory to define appropriate staffing levels as those that exceeded predefined rates within acceptable and tolerable limits (50% and 80%, respectively).⁶ That study found that work load was acceptable 42% of the time and tolerable 71% of the time. Appropriate staffing based on work load was achieved only 45% of the time.

Traditionally, nursing schedules are determined without knowing the imbalances in work load.⁷ Part of the challenge to improving work assignment may simply be an inability to track work load. In a related sense, nursing has struggled with creating standardized nursing intervention terminologies for use in electronic health records.⁸ While the respiratory care profession is just beginning to address this issue,⁹ respiratory care has a fully developed system for reporting treatment times,³ which is an invaluable resource for measuring and documenting work load and hence work rate.

A novel data-mining and simulation approach to evaluate patient-to-nurse assignments ahead of nursing shifts has been proposed by Sundaramoorthi et al.¹⁰ Our study applies a similar concept, with a different method, to respiratory care. To our knowledge, it is the first to propose the concept of work rate in respiratory care. We focused on the new variable, instantaneous work rate, instead of average work load as a potential new strategy by which to make RT staffing assignments. This approach has provided insights regarding the limits of current staffing models and opportunities to improve staffing to optimize timely delivery of RT services.

Our practice has been to assign each staff member 300 min of work in a 480-min (8-hour) shift. This assumed that 63% (300/480) of the work load was scheduled (currently ordered) and 37% was unscheduled or wasted (eg, travel time), based on time/motion studies at our institution. However, our current data suggest a larger portion of unscheduled work (45%). Therefore, individual work assignments should be based on an average scheduled work load, which is 264 min (55% of 480 min). Thus, if a therapist is assigned 264 cumulative minutes of scheduled work in particular areas of the hospital, it is understood that the therapist will also be responsible for the unscheduled work that arises in those areas. Given the high assignment-area variability in scheduled versus unscheduled work in various work-assignment areas, it is clear that work assignments should be based on actual daily data. Other main findings and implications of this preliminary analysis are that:

• Basing assignments on average work load leads to periods of unachievable work rate, potentially causing missed treatments and staff dissatisfaction.

• Given current technology and traditional work-assignment practices, the ability to meet work peaks is limited, but it appears that staggering treatment times can help.

• Fair work assignments based on average work load should differentiate between scheduled and unscheduled work.

Our spreadsheet model showed that shifting treatment start times improved the distribution of work rate throughout the day, but this measure alone will not guarantee that labor demand never outstrips supply. Other methods of reducing work load, particularly unnecessary work load, should be explored (eg, protocols driven by therapist-initiated patient assessment). Once such measures are in place there may be little gained from daily treatment-time adjustments. However, the tool still has utility if used to review retrospective data (ie, at the end rather the beginning of the day) to see the effects of both treatment time redistribution and actual unscheduled work. Such analyses may provide ideas for further refinement of assignment practices.

A product of this analysis was our spreadsheet work-assignment tool (Fig. 5), which is linked to the hospital's MediLinks database. The spreadsheet can be automatically populated with assignment-area work loads, allowing the RT supervisor to easily and efficiently assign work load in geographically proximate areas to achieve an appropriate average scheduled work load (264 min in the current case). This strategy greatly reduces the time required to assign work. For example, the assignments in Figure 5 were made in less than 5 min and required no special training or experience. One of us (RLC) had developed and used this tool successfully in another hospital, without a link to a work-load database. Instead, the tool was linked to another spreadsheet that was populated by asking staff from the previous shift to estimate the number of treatments of each type they expected for the next shift in each assignment area. Those treatment volumes were then automatically converted to work loads, using standard treatment times, and the resulting area work loads were automatically ported to the work-assignment tool.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Work-assignment tool implemented in a spreadsheet. Work load in each area is accumulated to achieve a scheduled work load of approximately 264 minutes. Areas with zero scheduled work load are nevertheless assigned to specific therapists so that any unscheduled work that arises will be accommodated. Note that the tool indicates that 69 minutes of scheduled work are still unassigned for work area H080. This indicates that the work in that area must be shared with another therapist.