Making Sense of the Data How to Prioritize
I have all this data, but I do not know what to do with it.” This is a direct quote from a nurse manager at a rehabilitation hospital several years ago. Have you ever felt this way? Nurse Managers are usually inundated with data on productivity, quality measures, census data and trends, fall rates, turnover rates, patient satisfaction scores and more. Increasingly, numbers are used by payers, hospital leaders and the public to make decisions about where, how and when to provide care. It is, therefore, of paramount importance that the nurse leader understand what the numbers mean, and how to prioritize.
The first thing a nurse leader must understand about numbers is the difference between raw numbers and rates. For example, a nurse leader at a rehabilitation unit in an acute care hospital, may be alarmed that they had six falls last month when they only had three the prior month. At first glance, the nurse leader can see that as a serious issue; there were twice as many this month as the prior month. Looking further however, due to a significant change in census, the number of falls in the first month is 8 per 1000 patient days while the number in the more recent month fall rate was 6 per 1000 patient days. As a note, patient days are determined by the daily census times the number of days in the month. For example if a unit had an average daily census of 50 in a month of 30 days, the total number of patient days is 1500. Since rehabilitation units often have a variable census, it is important to consider the impact of census on outcome data.
The second thing a nurse leader needs to understand is the use of comparative data and benchmarks. Patient satisfaction scores are often given by corporations who collect data from multiple sources and provide percentile rankings relevant to other clients of a particular vender. Using comparative data, a rehabilitation unit may decide that their scores are better than 50% of the clients for this survey company. Is that a good number? Much of that is determined by benchmarks. Benchmarks are established by the leaders of an industry. Typically, when we talk about benchmarks for patient satisfaction scores, the goal is better than 75 percent. So while the nurse manager may find that the patient satisfaction scores are better than 50 percent of clients, they still fall below the benchmark.
The third issue that the nurse leader needs to understand when it comes to numbers, is how to avoid making generalizations based on short term changes in data. For example, a nursing unit may find that the change on a patient assessment instrument has fallen by five points in the last month. The nurse leader may then decide that changes have to be made to educate the nursing staff on proper scoring, and recommend hiring new staff. Were other issues at work that affected this change in scores during this month? Perhaps this was a winter month and the unit accepted more patients with respiratory diagnoses than the prior month. It is well established that patients with a pulmonary diagnosis have scores on the patient assessment instrument that are less prone to dramatic swings from admission to discharge as opposed to patients admitted with traumatic brain injury. Just looking at one, or even a few months of data, can lead to misleading conclusions and ineffective decisions.
The fourth issue that the nursing leader must understand when looking at numbers, is the difference between correlation and causation. For example, a nursing leader, due to staffing vacancies, may hire multiple new employees. That same month, the nursing leader sees a spike in hospital acquired catheter-associated urinary tract infections (CAUTI). The nurse leader may decide that the new nursing staff caused the patients to have a higher rate of CAUTI and implement a new training program for all new employees. Without looking at all the data and possible causes behind the increased CAUTI, the nurse leader is making conclusions without establishing causation.
So once the nursing leader has the numbers, considered the rates and benchmarks, established a pattern and causation, what next? We’ll find out more in part 2 of this 3-part series.