what is the importance of variation to health-care organizations?
BMJ Qual Saf. 2011 April; 20(Suppl_1): i36–i40.
The meaning of variation to healthcare managers, clinical and health-services researchers, and individual patients
Duncan Neuhauser
iDepartment of Epidemiology and Biostatistics, Instance Western Reserve Academy, Cleveland, Ohio, USA
Lloyd Provost
2Assembly in Process Improvement, Austin, Texas, United states
Bo Bergman
iiiCentre for Health Comeback, Chalmers University of Technology, Gothenburg, Sweden
Abstract
Healthcare managers, clinical researchers and individual patients (and their physicians) manage variation differently to achieve different ends. First, managers are primarily concerned with the performance of care processes over time. Their time horizon is relatively short, and the improvements they are concerned with are pragmatic and 'holistic.' Their goal is to create processes that are stable and constructive. The analytical techniques of statistical process control effectively reflect these concerns. Second, clinical and health-services researchers are interested in the effectiveness of intendance and the generalisability of findings. They seek to command variation by their study design methods. Their primary question is: 'Does A cause B, everything else beingness equal?' Consequently, randomised controlled trials and regression models are the inquiry methods of option. The focus of this reductionist approach is on the 'average patient' in the group existence observed rather than the private patient working with the individual intendance provider. Third, private patients are primarily concerned with the nature and quality of their own care and clinical outcomes. They and their care providers are not primarily seeking to generalise beyond the unique individual. Nosotros propose that the gold standard for helping individual patients with chronic conditions should be longitudinal factorial design of trials with individual patients. Understanding how these three groups deal differently with variation can help appreciate these iii approaches.
Keywords: Control charts, bear witness-based medicine, quality of intendance, statistical procedure control
Introduction
Wellness managers, clinical researchers, and individual patients need to empathize and manage variation in healthcare processes in different time frames and in different ways. In short, they ask different questions nearly why and how healthcare processes and outcomes change (table i). Disruptive the needs of these three stakeholders results in misunderstanding.
Table 1
Meaning of variation to managers, researchers and individual patients: questions, methods and time frames
| Role | Question | Methods | Time frame | Variation |
| Wellness managers | Are we getting improve? | Control charts, holistic alter | Real time, months' variation | Creating stable processes, learning from special cause |
| Clinical and health-services researchers | Other things equal, does A cause B? | Randomised controlled trials, regression models; reductionist | Not urgent, years | Eliminate special-cause variation, examination for significance, focus on mean values |
| Private patient (and provider) | How can I get better? | Longitudinal, factorial designs | Days, weeks, lifelong | Help in understanding the many reasons for variation in health |
Health managers
Our all-encompassing experience in working with healthcare managers has taught us that their main goal is to maintain and meliorate the quality of care processes and outcomes for groups of patients. Ongoing care and its improvement are temporal, then in their state of affairs, learning from variation over time is essential. Data are organised over time to respond the fundamental management question: is care today every bit good equally or better than information technology was in the past, and how likely is it to exist better tomorrow? In answering that question, it becomes crucial to understand the difference between common-cause and special-crusade variation (as will be discussed later on). Common-cause variation appears as random variation in all measures from healthcare processes.1 Special-cause variation appears equally the effect of causes outside the core processes of the work. Management can reduce this variation by enabling the easy recognition of special-crusade variation and by changing healthcare processes—past supporting the use of clinical practice guidelines, for instance—simply common-cause variation can never exist eliminated.
The magnitude of common-crusade variation creates the upper and lower control limits in Shewhart command charts.2–5 Such charts summarise the work of health managers well. Figure 1 shows a Shewhart control chart (p-nautical chart) developed by a quality-comeback squad whose aim was to increase compliance with a new care protocol. The clinical records of eligible patients discharged (45–75 patients) were evaluated each week past the team, and records indicating that the complete protocol was followed were identified. The baseline control nautical chart showed a stable process with a centre line (average performance) of 38% compliance. The team analysed the aspects of the protocol that were not followed and developed process changes to make information technology easier to consummate these particular tasks. Afterward successfully adapting the changes to the local environs (indicated past weekly points above the upper command limit in the 'Implementing Changes' catamenia), the team formally implemented the changes in each unit. The squad connected to monitor the procedure and somewhen developed updated limits for the chart. The updated chart indicated a stable process averaging 83%.
Annotated Shewhart command nautical chart—using protocol.
This control nautical chart makes information technology clear that a stable only inferior process was operating for the starting time 11 weeks and, by inference, probably before that. The annotated changes (testing, adapting and implementing new processes of care) are linked to designed tests of change which are special (assignable) causes of variation, in this case, to improvement after calendar week 15, afterward which a new better stable process has taken hold. Notation that in that location is common-cause (random) variation in both the old and improved processes.
After updating the control limits, the chart reveals a new stable process with no special-crusade variation, which is to say, no points in a higher place or below the control limits (the dotted lines). Note that the change subsequently week xv cannot easily be explained by chance (random, or common-crusade, variation), since the probability of 13 points in a row occurring past chance higher up the baseline control limit is one divided past 2 to the 13th ability. This is the aforementioned likelihood that in flipping a money 13 times, it will come up up heads every fourth dimension. This level of statistical power to exclude randomness as an explanation is not to be found in randomised controlled trials (RCTs). Although there is no hard-and-fast rule about the number of observations over time needed to demonstrate process stability and establish change, we believe a persuasive control chart requires 20–thirty or more observations.
The managing director'southward task demonstrates several of import characteristics. Start is the need to ascertain the central quality characteristics, and choose amongst them for focused improvement efforts. The option should be made based on the needs of patients and families. The importance of these quality characteristics to those being served ways that speed in learning and improvement is important. Indeed, for the healthcare director, information for improvement must exist as rapid every bit possible (in real time). Twelvemonth-sometime research data are not very helpful hither; just-in-fourth dimension performance data in the easily of the decision-makers provide a potent opportunity for rapid comeback.6
Second, managerial modify is holistic; that is, every chemical element of an intervention that might help to improve and can be done is put to use, sometimes incrementally, but simultaneously if need be. Healthcare managers are actively working to promote measurement of process and clinical outcomes, accept problems in organisational performance seriously, consider the root causes of those bug, encourage the germination of problem solving clinical micro-system teams and promote the use of multiple, evolving Plan–Do–Study–Act (PDSA) tests of change.
This kind of improvement reasoning can be applied to a wide range of care processes, large and modest. For case, good surgery is the appropriate combination of hundreds of private tasks, many of which could be improved in small ways. Aggregating these many smaller changes may result in important, appreciable improvement over fourth dimension. The protocol-driven, randomised trial inquiry approach is a powerful tool for establishing efficacy but has limitations for evaluating and improving such complex processes as surgery, which are continually and purposefully irresolute over time. The realities of clinical comeback call for a motion from after-the-fact quality inspection to building quality measures into medical data systems, thereby creating real-time quality data for providers to human activity upon. Caring for populations of similar patients in similar ways (economies of calibration) can be of particular value, because the resulting big numbers and procedure stability tin can assist speedily demonstrate variation in care processesvii; very tight command limits (minimal common-cause variation) allow special-cause variation to be detected more quickly.
Clinical and health-services researchers
While quality-direction thinking tends towards the use of data plotted over fourth dimension in control-chart format, clinical researchers think in terms of true experimental methods, such as RCTs. Health-services researchers, in contrast, think in terms of regression analysis as their primary tool for discovering explainable variation in processes and outcomes of care. The data that both communities of researchers employ are mostly collected during stock-still periods of fourth dimension, or combined beyond time periods; neither is usually concerned with the analysis of data over time.
Take, for example, the question of whether age and sex are associated with the ability to undertake early ambulation after hip surgery. Clinical researchers try to command for such variables through the apply of entry criteria into a trial, and random assignment of patients to experimental or command group. The usual health-services research approach would be to employ a regression model to predict the event (early ambulation), over hundreds of patients using historic period and sex as independent variables. Such research could show that age and sexual practice predict outcomes and are statistically significant, and that peradventure 10% of the variance is explained by these two contained variables. In contrast, quality-improvement thinking is likely to conclude that 90% of the variance is unexplained and could be common-cause variation. The health-services researcher is therefore likely to conclude that if we measured more variables, we could explain more of this variance, while comeback scientists are more likely to conclude that this unexplained variance is a reflection of mutual-crusade variation in a good process that is nether control.
The entry criteria into RCTs are carefully defined, which makes it a claiming to generalise the results beyond the kinds of patients included in such studies. Restricted patient entry criteria are imposed to reduce variation in outcomes unrelated to the experimental intervention. RCTs focus on the departure between betoken estimates of outcomes for entire groups (control and experimental), using statistical tests of significance to show that differences between the two arms of a trial are not likely to be due to chance.
Individual patients and their healthcare providers
The question an individual patient asks is different from those asked by manager and researcher, namely 'How can I get better?' The answer is unique to each patient; the question does not focus on generalising results across this person. At the same time, the question the patient's dr. is asking is whether the grouping results from the best clinical trials will utilize in this patient's case. This question calls for a different inferential approach.8–10 The cost of projecting full general findings to private patients could be substantial, every bit described below.
Consider the implications of a drug trial in which 100 patients taking a new drug and 100 patients taking a placebo are reported as successful because 25 drug takers improved compared with 10 controls. This difference is shown as not likely to be due to chance. (The drug company undertakes a multimillion dollar advertising entrada to promote this breakthrough.) However, on closer examination, the meaning of these results for private patients is non so articulate. To begin with, 75 of the patients who took the drug did not benefit. And amongst those 25 who benefited, some, perhaps 15, responded extremely well, while the size of the benefit in the other 10 was much smaller. To take simply the fifteen 'maximum responders' have this drug instead of all 100 could save the healthcare system 85% of the drug'due south costs (as well every bit reduce the chance of unnecessary adverse drug effects); those 'savings' would, of grade, also reduce the drug visitor'southward sales proportionally. These considerations brand information technology clear that looking at more than than group results could potentially make an enormous difference in the value of enquiry studies, particularly from the point of view of individual patients and their providers.
In light of the above concerns, we advise that the longitudinal factorial written report design should be the gold standard of evidence for efficacy, particularly for assessing whether interventions whose efficacy has been established through controlled trials are constructive in individual patients for whom they might exist appropriate (box one). Take the case of a patient with hypertension who measures her blood pressure at least twice every twenty-four hours and plots these numbers on a run nautical chart. Through this informal observation, she has learnt about several factors that result in the variation in her blood pressure readings: time of day, the 3 different hypertension medicines she takes (not always regularly), her stress level, eating salty French fries, exercise, meditation (and, in her case, saying the rosary), and whether she slept well the night earlier. Some of these factors she tin can control; some are out of her control.
Since she is accustomed to monitoring her blood pressure over time, she is in an splendid position to carry out an experiment that would assistance her optimise the effects of these diverse influences on her hypertension. Working with her primary intendance provider, she could, for example, set a table of randomly chosen dates to make each of several of these changes each day, thereby creating a systematically predetermined mix of these controllable factors over fourth dimension. This factorial design allows her to measure the effects of private inputs on her claret force per unit area, and even interactions among them. After an appropriate number of days (perhaps 30 days, depending on the trade-off between urgency and statistical power), she might conclude that one of her 3 medications has no effect on her hypertension, and she can stop using it. She might likewise detect that the combination of exercise and consistently low common salt intake is as effective as either of the other 2 drugs. Her answers could well exist unique to her. Planned experimental interventions involving single patients are known as 'N-of-1' trials, and hundreds have been reported.ten Although longitudinal factorial pattern of experiments has long been used in quality engineering, every bit of 2005 there appears to have been only one published instance of its use for an individual patient.8 nine This method of investigation could potentially get widely used in the future to establish the efficacy of specific drugs for individual patients,11 and perchance even required, particularly for very expensive drug therapies for chronic conditions. Such individual trial results could exist combined to obtain generalised noesis.
This method tin exist used to evidence (1) the contained consequence of each input on the outcome, (2) the interaction effect between the inputs (perhaps neither drug A or B is effective on its ain, but in combination they work well), (iii) the outcome of different drug dosages and (iv) the lag time betwixt treatment and result. This approach will not be practical if the outcome of interest occurs years later. This method volition be more practical with patient access to their medical record where they could monitor all v of Bergman's core health processes.12
Agreement variation is one of the cornerstones of the scientific discipline of improvement
This broad understanding of variation, which is based on the work of Walter Shewart in the 1920s, goes well beyond such simple bug equally making an intended departure from a guideline or recognising a meaningful change in the consequence of care. Information technology encompasses more than good or bad variation (coming together a target). Information technology is concerned with more than than the variation found by researchers in random samples from big populations.
Everything we observe or measure varies. Some variation in healthcare is desirable, fifty-fifty essential, since each patient is unlike and should be cared for uniquely. New and improve treatments, and improvements in care processes result in beneficial variation. Special-crusade variation should lead to learning. The 'Plan–Practise–Study' portion of the Shewhart PDSA cycle can promote valuable change.
The 'act' stride in the PDSA cycle represents the arrival of stability after a successful comeback has been made. Reducing unintended, and specially harmful, variation is therefore a key comeback strategy. The more variation is controlled, the easier it is to notice changes that are not explained by risk. Stated differently, narrow limits on a Shewhart command chart make information technology easier and quicker to detect, and therefore respond to, special-cause variation.
The goal of statistical thinking in quality improvement is to brand the available statistical tools as simple and useful as possible in meeting the primary goal, which is non mathematical correctness, but comeback in both the processes and outcomes of care. It is not fruitful to enquire whether statistical process command, RCTs, regression equations or longitudinal factorial design of experiments is best in some absolute sense. Each is advisable for answering dissimilar questions.
Forces driving this new way of thinking
The idea of reducing unwanted variation in healthcare represents a major shift in thinking, and it volition have fourth dimension to be accustomed. Forces for this change include the computerisation of medical records leading to public reporting of intendance and outcome comparisons between providers and around the world. This in plow will promote pay for performance, and preferred provider contracting based on guideline utilise and practiced outcomes. This manner of thinking well-nigh variation could spread beyond all five core systems of health,12 including self-intendance and processes of good for you living.
Footnotes
Competing interests: None.
Provenance and peer review: Not commissioned; externally peer reviewed.
References
i. Lighter DE. Advanced Performance Improvement in Wellness Intendance. Sudbury, MA: Jones and Bartlet, 2011, Chapter five. [Google Scholar]
two. Bergman B, Klefjso B. Quality from Customer Needs to Client Satisfaction. 3rd edn Lund, Sweden: Studentlitteratur, 2010 [Google Scholar]
3. Berwick D. Controlling variation in health care: a consultation with Walter Shewhart. Med Care 1991;29:1212–25 [PubMed] [Google Scholar]
4. Moen RM, Nolan TW, Provost LP. Quality Comeback through Planned Experimentation. 2nd edn New York: McGraw-Loma, Inc; 1998 [Google Scholar]
5. Nolan TW, Provost LP. Understanding Variation, Quality Progress. ASQC, 1990 [Google Scholar]
half dozen. Lifvergren Due south, Gremyer I, Hellstrom A, et al. Lessons from Sweden's First Big Scale Implementation of Six Sigma in Health Intendance. Oper Manage Res 2010;3:117–28 [Google Scholar]
7. Diaz M, Neuhauser D. Pasteur and parachutes: when statistical process command is meliorate than a randomized clinical trial. Qual Saf Health Care 2005;14:140–iii [PMC costless article] [PubMed] [Google Scholar]
viii. Thor J. Getting Going on Getting Better: How is Systematic Quality Improvement Estalished in a Health Care Organization—Implications for Alter Direction and Theory. Doctoral dissertation. Stockholm: Medical Management Eye, Karolinska Institute, 2007. and summarized in Quality and Safety in Health Care. [Google Scholar]
9. Olsson J, Terris D, Elg M, et al. The one person randomized controlled trial. Qual Manag Wellness Care 2005;fourteen:206–16 [PubMed] [Google Scholar]
11. Hebert C, Neuhauser D. Improving hypertension care with patient generated run charts. Qual Manag Wellness Care 2004;thirteen:174–7 [PubMed] [Google Scholar]
12. Bergman B, Neuhauser D, Provost 50. Five main processes in healthcare—a citizen perspective. BMJ Qual Saf 2011;20(S1):i41–i42 [PMC gratis commodity] [PubMed] [Google Scholar]
Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3066789/#:~:text=Some%20variation%20in%20healthcare%20is,variation%20should%20lead%20to%20learning.