Variation is the way the performance of certain processes changes over time. This can happen to the process at any time, say on a day-to-day basis, week-to-week basis or even a month-to-month basis. The changes can be natural or at times just random, and occur in all processes so it is important that they are anticipated. These changes are always a result of the elements that affect the process such as the materials used, procedures, equipment or electronic systems among more. Therefore, understanding the nature of variations in healthcare is vital in the decision making process on improvement efforts.
In the healthcare industry, unwarranted variations in medical practice are often costly and sometimes even deadly. You may find that an approach towards a patient’s case in one town is a major surgery, while in another town it's watchful waiting. Some analysis done on the Medicare data showed that per-capita spending for any enrollee in Miami was almost two and a half times greater than that in Minneapolis, considering that the data had even already been adjusted for sex, age and race. And worse still, according to a report by the National Committee for Quality Assurance, over 55,000 lives are lost annually because physicians don’t use evidence-based medicine to steer their care delivery.
Therefore, managing variation is crucial to quality improvement and all concerned parties should be well aware of how to go about it. Quality improvement is majorly concerned with two aspects or types of variation namely;
- Common cause variation, and
- Special-cause variation
Common cause variation refers to the random deviations or fluctuations present in a stable healthcare process while special-cause variation refers to the unpredictable deviations resulting from external factors.
The approach used in managing variations in healthcare depends on the goals, priorities, and perspectives of the improvement leader. We will take a look at three different perspectives from which variation can be managed for improvement efforts. These perspectives include the approach from a clinical researcher’s angle, healthcare manager’s angle, and an individual patient’s angle. All these parties have different goals, methodological approaches and time horizons on managing variation.
Clinical researcher perspective
Traditional randomized controlled trials (RCTs) were designed to establish treatment or intervention efficiency in a certain specified population while holding all other factors constant. Here, the researchers aimed at maximizing internal validity. Researchers use randomization to classify patients, organizations, and clinicians into experimental and control groups in their study on healthcare variations. They may also use methods such as stratification to increase their understanding of variation.
Traditional research methods are faced with many challenges such as limited study populations, inability to evaluate complex processes, and study populations not being truly representative of the real world. This makes the applicability of their findings in clinical care unclear. However, using alternative modern designs like the hybrid, researchers can now make their findings clearer and better help manage variations.
Healthcare manager perspective
Healthcare managers seek to study and better manage variations in patient populations by observing various processes and their outcomes. They use real-time data to study and compare the past, present and desired performance in order to reduce any undesired variations and strengthen the desired variations. Plus, the healthcare managers implement best care practices and use them as a point of reference when it comes to performance.
From the patient perspective, an individual patient is concerned with how a particular treatment will enable them to achieve healthcare outcomes that are similar to those established in study populations. That is, is there going to be a significant variation or not? Although research findings by RCTs create the basis for evidence-based care practice and managers use these findings to help them with population management, they offer less guidance on how a single patient can achieve the average benefits monitored across a population of patients.
However, with different methodological approaches, patient and providers can be able to systematically evaluate the effects of various disease management variables on individual patient health outcomes. This enables both providers and patients to collect and analyze data in real time so as to improve individual patient outcomes.
Variations in Healthcare and Risk Stratification
Risk stratification is used to predict, prioritize and prevent.
When we say "prediction" we mean predict what might happen in the future for each patient. Predict who may or may not be readmitted, may need higher levels of care, or who is at risk for their diseases or symptoms to get worse. We want to predict anything that may or may not happen to a patient before it occurs. Once future outcomes of patients are predicted, you can prioritize the needs of each patient, based on these predictions. Hopefully, by prioritizing more at-risk patients, you can stop or slow the progression of an illness or disease with certain interventions. Once a patient is prioritized to get the care he or she needs, pre-emptively, we hope to prevent future re-admissions, symptoms getting worse, or, worst case scenario, death.
Knowing that someone has stage 1 diabetes, smokes and is overweight, it would be accurate to predict that this person’s diabetes may get worse without some lifestyle changes, as well as counseling and maybe even some initial treatments or medicines. Because of this person’s personal factors, you have predicted their diabetes can and likely will get worse. From there, you can prioritize this patient.