There are so many forms of data used where I work, that it is practically one way of describing what it is we do. Nursing and operating a hospital are so full of different uses of data that it is a challenge to break them down into individual uses. There is data used for budgeting, staffing, and supplies. Data is collected on demographics of the patient population. Data is measured constantly in patients; vital signs, lab values, and measurements of all sorts. We also conduct clinical trials in vast numbers collecting data on efficacy, treatment variation, modalities, and outcomes.
2. One example of descriptive statistics we use in our work place is yearly charts depicting prevalence of falls hospital wide. These charts are shown in a number of ways and are divided by unit so that each unit can be compared one with the other. Additionally, they have charts showing the yearly trends by unit and hospital-wide. They get fairly detailed on some of them showing specific categories of fall occurrence, for example; assisted vs. unassisted; Falls due to toileting; falls by time of day; falls by assessed risk category. This practice is highly effective and allows everyone to see where, when, and how falls are occurring. It also provides insight into how to go about reducing the instances.
3. Where I work we do use inferential statistics, in fact we apply it a lot. We administer chemotherapy and study its effects. We collect data from sample populations (trials) and based on the analysis develop protocols to treat patients for side effects that are commonly associated with the treatment, regardless of actual occurrence or rather before assumed occurrence. An example of this is the chemotherapy “nadir” effect; data that shows a drop in blood cell count in ten to fourteen days after initiation of therapy. We have a slew of measures to prevent this drop in blood counts including a drug we developed in-house; GCSF (Injected daily to induce production of blood cells usually for five days). Other measures are also instituted to prevent infection should blood cell counts drop, such as instructing patients on eating low microbial diets. We give other medications to prevent known side effects like nausea, mouth sores, and loss of appetite. In a sense the use of inferential statistics allows us to be proactive rather than reactive and hence more effective.
4. Data at the nominal level is used extensively. Patients are grouped by the type of cancer they have and assigned to the unit that specializes in dealing with that diagnosis. They are then assigned a room based on gender.
Ordinal level data is used when assigning patients to a room. Those assessed as higher risk for falls are kept closer to the station whereas those that are low risk are placed father away. The highest risk patients are assigned to rooms with a companion to monitor them continuously.
Data at the interval level/ratio level data is very common in the hospital setting anything from age, to measures of pressures, temperatures, blood counts, amounts of fluids in and out and scores of other measurements.
5. Statistical analysis is used in many ways. As stated before in order to prevent some effects of drugs. It is used to budget each unit and the hospital as a whole. It is used to devise ways of increasing productivity, patient safety, and desirable outcomes. The use of statistics allows us to evaluate what has worked and why in reducing falls and allows us to apply good working models to other areas. Analyzing statistics plays a role in staffing. It tells us what the highest traffic days of the week are likely and allows us to staff accordingly. Likewise, for the time of the year there are predictable increases or decreases in patient levels. At the same token it can show us a clear correlation between staffing numbers and incidences of errors. We have an entire online database (anonymous) for reporting any sort of error or near miss, all of which are studied to improve quality of care. Most of all, we remain a research hospital. We run huge numbers of studies and clinical trials that are used to develop treatments, protocols, and new medicines.