The Quality-Cost Connection: How to evaluate your staffing decisions

Staffing effectiveness standards take effect July 1

By Patrice Spath, RHIT
Brown-Spath Associates
Forest Grove, OR

Is patient safety threatened when there are fewer registered nurses at the bedside? When the workload of the respiratory therapy staff goes up, are more treatments missed? Is there an increased incidence of pulmonary complications in critical care patients when staff work more overtime hours? It is these types of questions that hospitals are expected to address after implementing the new staffing effectiveness standards of the Joint Commission on Accreditation of Healthcare Organizations (JCAHO). The standards, scheduled for implementation on July 1, 2002, are designed to improve an organization’s understanding of the relationship between staffing and outcomes. The goal is to ensure that staffing is adequate to protect patients from unintended harm and maintain high satisfaction levels.

The staffing effectiveness standards do not replace current human resource management requirements. It still is necessary for the hospital to have a plan for providing adequate and capable staff and a method for assessing staff competencies. What’s being added is a requirement that the organization’s leaders use information to evaluate the effectiveness of staffing decisions in terms of outcomes. Quality patient care is dependent on adequate numbers of capable staff who can meet the complex needs of patients. To evaluate how well the hospital is meeting this goal, the Joint Commission expects the organization to use several indicators to identify potential staffing problems. Data should be gathered that enable both the staff and the senior leaders to collaboratively monitor the relationship between quality patient outcomes and staffing and make changes when staffing decisions are negatively impacting expectations.

The new staffing effectiveness standards have no phase-in period. If your hospital is scheduled for a Joint Commission survey in September 2002, surveyors will expect to see a 2-month track record of data collection and analysis. Hospitals surveyed in January 2003 will be expected to have a 6-month track record. A 12-month track record is expected for hospitals not scheduled for survey until July 2003 or beyond.

Hospitals must select indicators or measures that can be used to evaluate staffing effectiveness. The organization as a whole must identify two human resource variables and two clinical outcome or service variables that will help the organization evaluate the impact of staffing decisions. The standards require that at least one of the human resource indicators and one of the clinical or service indicators be chosen from the list provided in the standards manual.

It may be best to start with your "hunches" about possible cause-and-effect relationships rather than merely picking all four indicators from the list provided by the Joint Commission. Examples: "I wonder if the number of overtime hours affects the number of medication errors? Is there a relationship between the supervisor-to-staff ratio and the number of postoperative complications? Is there a relationship between test turnaround times in the emergency department and patients who left without being seen?" Ask staff to provide their input. They very likely have some good hunches worth checking out.

Your organization must have a reasonable explanation for choosing indicators. Ideally, senior leaders select measures that are most apt to tell them what they want to know about the effect of staffing decisions on important clinical or service outcomes. Don’t shy away from studying a significant issue just because the indicators are not on the JCAHO list or data are not currently being gathered for the measure. At the time of survey, you’ll need to explain how the indicators you’ve chosen relate to your organization’s improvement priorities and patient population.

Once the indicators have been selected, it’s time to identify which departments are affected by the measures. Although research studies of staffing effectiveness primarily have focused on nursing staffing, the JCAHO standards apply to all departments that provide direct patient care (for example, nursing, respiratory therapy, physical therapy, social services, and chaplain) and indirect caregivers (for example, pharmacy, laboratory, housekeeping, and radiology). Suppose your organization is interested in evaluating a "hunch" about the cause-and-effect relationship between productive work hours and patient falls. Departments affected by these measures would include nursing, physical therapy, and perhaps even housekeeping and volunteer services. One variable (falls) would be compared to the other variable (productive work hours) for all the affected departments.

The results can be reported by month or by quarter, depending on the volume of data. The information should be displayed in a manner that allows people to see the relationship between the variables and performance expectations. (For an example of a matrix report illustrating the relationship between three human resources measures and three clinical measures, see chart, below.)

Matrix Report of Staffing Effectiveness Indicators

Expected Performance Last 12
Jan Feb  Mar Apr May Jun
# falls per 100 patient days < 2.0 2.25  1.3  2.0 1.4 1.0 2.1 1.8
# medication errors per 100 patient days < 40.0   42.0 44.0 37.0  28.0 33.0  31.0  36.0
% of patients restrained < 25% 27% 18%  32% 23% 21% 22% 26%
% of positions filled by agency staff < 20% 15% 23%  18% 20% 15% 18%  18%
% work hours recorded as overtime < 10% 7% 18% 15%  8% 9% 8% 12%
% vacant staff positions < 20% 20% 29% 25% 20% 18% 20%  15%

Scatter diagrams can be a useful way to analyze the relationship between two variables. People can see the direction of the relationship (positive, negative, etc.) and the strength of the relationship. As more data are gathered, new data points are added to the diagram. A positive relationship is indicated by an ellipse of points that slopes upward, demonstrating that an increase in the cause variable also increases the effect variable. A negative relationship is indicated by an ellipse of points that slopes downward, demonstrating that an increase in the cause variable results in a decrease in the effect variable.

A diagram with a cluster of points such that it is difficult or impossible to determine whether the trend is upward-sloping or downward-sloping indicates that there is no relationship between the two variables.

The scatter diagram (to see diagram, click here) shows a positive relationship between test turnaround times in the emergency department (the cause) and number of patients who leave without been seen (the effect). Each data point represents the results from one month. There are 20 months of data displayed on the diagram.

The measurement aspect of the staffing effectiveness standards is somewhat like correlational research in which variables are measured so that an association between some set of variables can be identified.

If a relationship is found to exist, further investigation should be undertaken. Because data from correlational research cannot conclusively prove causality, the measures used to evaluate staffing effectiveness are considered "screening indicators." Positive correlations are merely an indication that further inquiry is needed, not that staffing practices immediately need to be altered.

The group responsible for reviewing the indicators, such as the administrative council, should "drill down" on any data that vary from the expected results or reveal a positive and undesirable correlation. The drill-down process is much like a root-cause analysis; ask "why, why, why" when evaluating staffing-related issues. For example, reduced staffing levels may appear to be related to an increase in patient complaints. The solution given is to increase staffing. This may be very appropriate as one short-term option, but may miss the bigger picture. Why are patients dissatisfied? Why are expectations not being met? What other factors contributed to patient dissatisfaction and why? By answering the "drill-down" why questions, the result may lead to a better process and sustained improvement in patient satisfaction.

If performance expectations are met, no relation exists between staffing indicators and clinical or service indicators, and you are satisfied that the results are statistically valid, then change the variables you are investigating. Don’t keep measuring the same indicators to evaluate staffing effectiveness. Check out other "hunches" to determine if your organization’s staffing decisions are safe and adequate to provide patients with needed services.