Quality and Safety in Imaging

© Springer International Publishing AG 2018

Lluís Donoso-Bach and Giles W. L. Boland (eds.)Quality and Safety in ImagingMedical Radiologyhttps://doi.org/10.1007/174_2017_145

Framing the Issues

Giles W.L. Boland¹, ²  

(1)

Department of Radiology, Brigham and Women’s Hospital, Francis Street, Boston, MA, USA

(2)

Harvard Medical School, Boston, MA, USA

Giles W.L. Boland

Email: [email protected]

Quality and safety is increasingly ascendant in medicine as systems are focused on delivering better value and outcomes to patients and payers. However, even with the many checks and balances being introduced into clinical workflows, medical systems are still challenged to deliver consistent, evidenced-based best practices at the point of care. In the United States, the Institute of Medicine believes that close to 100,000 deaths/annum are created by medical error and some opinions believe that number to be closer to 400,000 lives. There is no known worldwide statistic but it almost certainly runs into the millions. While radiological procedures may usually seem non-life threatening, there is still considerable risk, real or perceived. Certainly, invasive interventional procedures do carry significant risk, even death (i.e., angiography or percutaneous biopsy). Other procedures have theoretical risk such as the effects of radiation dose exposure (even at lower doses) mainly from Computed Tomography, discussed elsewhere in this book. Furthermore, there is widespread variation in the use of appropriate examinations (imaging tests) for a particular condition, sometimes referred to as appropriateness (see other chapters in this book). In fact, despite much evidence on the use of appropriate best practices for radiological procedures (usually promulgated by national radiological societies), variation in the practice of radiology abounds, usually with no two departments alike delivering similar practices and operating procedures. What might seem appropriate in one department is often not seen in another—for instance, what is viewed as an acceptable radiation dose varies across regions, towns, and sometimes even within the same health organization. Given this widespread variation, legislative, payer, and professional bodies are now finding this scenario unacceptable and are introducing legislation or pay-for-performance measures to drive organizations to deliver more consistent and better care with outcomes that meet certain predetermined standards. Furthermore, patients themselves are now demanding better outcomes and less variation, particularly as it has become more evident from the press that outcomes can significantly vary from one organization to another. This has come at a time when demand for imaging services is busier than ever as referrers continue to see imaging as a key tool to reach a diagnosis earlier, monitor therapy more closely, and/or cure and palliate patients through innovative interventional therapies. This significant increase in radiological volume has sometimes come at the cost of quality (and even safety) as radiologists and departments are busier than ever trying to keep up with demand of simply performing and interpreting the procedures. Often departments are just too inundated with the workload to take a step back to rethink fundamentally how quality and safety initiatives can be reorganized in a meaningful and systematic way to drive the delivery of care towards better practices and outcomes. Quality and safety measures, which are often difficult to measure, let alone deemed as meaningful in the first place, are then sometimes seen as an afterthought. Even experts often struggle to define standards and then agree upon them. Furthermore, measures put in place to monitor quality and safety are frequently imposed from afar, often by payers (i.e., large bureaucracies such as the Center for Medicare and Medicare Service or the National Health Service) and therefore deemed onerous and unnecessarily imposing by front-line providers. This can result in frustration and ambivalence towards the quality and safety agenda. The approach of many radiologists to many of the quality and safety measures is to simply check the box so they can either meet their mandatory compliance standards or, in increasing circumstances, actually get paid. There is a common belief that many of the quality and safety standards are either only tangentially relevant or sometimes not meaningful at all. Added to this frustration, the practice of medicine and radiology keeps changing and even experts find it difficult to keep up with new technologies, treatments, and new care pathways such that creating meaningful, up-to-date, and relevant metrics inevitably lags the innovation. Finally, although pay-for-performance measures are now tying part of payments to performance (sometimes quality and safety), much of what can be achieved through quality and safety initiatives is not reimbursed. Considering the numerous other non-remunerated regulatory and compliance measures required from radiologists, quality and safety initiatives are often viewed as overly burdensome and are relegated to the domain of just doing the right thing for the patient rather than a compelling reason to do so.

Despite these challenges and the increasing nonclinical workload that radiologists are facing, it is imperative that all caregivers and departments develop, implement, and monitor a robust quality and safety program to remove unnecessary variation, deliver best practices at the point of care, and ultimately deliver better outcomes for their patients. Achieving this requires a cultural shift within the organization, sometimes referred to as the culture of safety. This starts with leadership whose role is to impart a compelling reason to their staff as to why quality and safety is integral to every aspect of the workflow, why measurement is important, and why change is mandatory when standards do not meet best practices. In other words, the work needed is not optional and the programs and people put in place need to be held accountable to the mission and goal at hand. As with all leadership, effective translation of the vision will require choosing the right teams to develop meaningful strategies, tactics, and tools to deliver better quality and that these teams need to work with the wider department to ensure consistent delivery of the solutions. Constant monitoring, feedback, and sharing of the data will be necessary to iterate and improve as well as benchmarking departmental and individual performance. This cultural pivot often takes years to implement and requires constant vigilance to ensure that teams and individuals do not lose sight of the primary goal of the processes—the delivery of better outcomes for patients. Otherwise commitment can quickly unravel and quality and safety will again be viewed as a burdensome and relatively unnecessary part of their workload. Performance monitoring and measurement is critical to driving cultural change and offers managers the opportunity to transform their departments towards better practices. Fortunately, while still challenging, measurement is becoming more seamless through electronic health records and data capture and display, which is more presentable and understandable and importantly, up to date. Departments are developing scorecards and dashboards to help providers understand their performance either instantly as in the use of dashboards (i.e., how many patients have been cleared for MRI safety checks on a given day) or over several weeks or months as in scorecards that look at trends in performance over a period of time. Both are meaningful tools with which to benchmark quality and safety practices from which teams can then determine if further improvement and change is needed.

It cannot be understated how important a pivot to a culture of safety must be sustained, because it changes the mindset of a department to gear all aspects of the operations with quality and safety in mind. Rather than quality and safety being seen as some arbitrary and unnecessary imposition, it fully embraces the Hippocratic oath of first do no harm, a foundational medical doctrine that goes as far back as, well… Hippocrates. While this oath is rarely formally taken by physicians nowadays, it surely is in the hearts and minds of all providers as they strive to do the best they can for patients. Yet, doing one’s best is often not sufficient, it is better to know what to do and then do one’s best (attributed to Edwards Demings). In other words, while caregivers honestly strive to do their best for patients, in reality it may fall well short of current best practices either because of lack of knowledge or systems that are not in place to aid providers to deliver care of the highest quality and safety. Leaders have the challenging task of placing quality and safety at the core of their operations which can then drive all departmental practices accordingly and in a manner that all staff can embrace and support. Only when every member of the overall team believes that working together in a data-driven, supportive, non-punitive, and transparent framework will departments approach the culture of safety.

Once leadership provides the vision for the quality and safety agenda, it is advantageous to frame the approach by considering the operations in totality rather than piecemeal, where individual activities are not viewed as connected or integrated into a larger framework. For instance, a quality and safety agenda may do sterling work on reducing radiation dose for specific CT procedures and the department may be led to believe that they are excelling in this particular arena. While reduction of CT dose is unquestionably appropriate and necessary, the quality efforts to achieve this can quickly be undermined or even rendered useless depending on patient circumstances. For instance, no matter how much work has been invested into reducing dose and no matter how low a CT dose has been achieved, it is meaningless to those patients who underwent a CT, which was not indicated in the first place. In other words, the efforts to reduce CT dose have not been tied to the necessary efforts to reduce imaging inappropriateness. Similarly, efforts by radiologists to become more subspecialized towards precision reporting will be undermined if the report they are generating for an examination was for a test that was inappropriate or non-indicated. In that sense, all unnecessary activities can be viewed as waste and ultimately error, the antithesis of quality and safety. Similarly, variation in performance can also be considered as waste and error as best practice standards are not being consistently met at the point of care. Increasingly waste and variation are being viewed as a cost to the overall system, a major driver for inefficiency in health care (not just one of morbidity and mortality). The Institute of Medicine in the United States believes that medical inefficiencies (waste) contribute up to 33% of medical costs (over $1 trillion in annual waste) so quality and safety measures are now considered a critical component of reducing waste and costs in the system. Furthermore, cost can be understood as not just financial. Redundant and inappropriate care can lead to unnecessary anxiety for patients and inappropriate use of their time and other resources.

So quality and safety measures are now a central and major focus of policy makers, payers, hospital leaders, patient advocates, and in turn care providers as they strive not just to reduce morbidity and mortality for patients but a whole host of other cost issues. Radiology services must in turn address these forces and acknowledge that the efforts ought to be comprehensive and overarching—and address every aspect of the radiology operations. To achieve this, leaders and managers must recognize that all radiology activities and operations are ultimately interlinked. Business leaders have recognized this for decades and some have used a value chain as a metaphor to help understand and frame their operations to improve performance, quality, and even safety. They teach that each component of an operation or workflow contributes to the overall performance of that operation, whether it be a service or a product. This metaphor is just as apt for radiology and it could be helpful to view the radiology operations as an imaging value chain and the delivery of best practices is only as strong as the weakest link in that chain. Therefore improving quality and safety in one domain (or link) does not necessarily translate into overall effectiveness if other up or downstream efforts have not been similarly addressed.

The imaging value chain can be simplistically imagined as the workflow from when a referrer orders an imaging test to when he or she receives a report, hopefully one that is actionable (see graphic). Hopefully the referring physician is familiar with the right test to order for the patient (image appropriateness) but not infrequently they do not know precisely the best test to order at that time for that patient with their current presenting complaint. This can be termed imaging inappropriateness, with some believing this could account for up to 30% of all imaging requests. Once the test is ordered it then needs to be scheduled and protocoled. Then the patient arrives at the imaging suite and the procedure is performed (could be either diagnostic or interventional). The images are acquired and transmitted (and stored) at which point the radiologist interprets the images (with increasing access to collateral biomarker data residing in electronic health record databases). The reports are then communicated to the referring physician, ideally actionable (meaning they are succinct, structured, precise, unambiguous, directional)—in other words a report that the referrer can then use to determine the next best course of action without unnecessary additional tests or actions which might only lead to additional waste and cost in the system.

Using the imaging value chain metaphor, it helps departments to view the operations as a whole and approach quality and safety initiatives as a systems approach so benefits in one part of the system can effectively be translated through to other parts. As discussed, too often managers do not envisage their operations holistically when devising quality and safety measures, rather efforts are fragmented and uncoordinated. For imaging appropriateness, tests should only be ordered when they are deemed absolutely necessary. Given the complexity and pace of modern medicine, this can only realistically be achieved through computerized decision support systems that guide referrers to order the right test for any given clinical scenario. This then sets the stage for the delivery of an actionable report downstream (as an inappropriate test is, by definition, non-actionable). Once the test is chosen, it behooves the operations to perform the test as quickly as possible (otherwise why would the test be necessary). This means scheduling the test expeditiously, ideally through sophisticated electronic order entry systems that allow referrers and patients to choose an imaging location of their choice and convenient time. This may not seem a quality measure but from a customer’s point of view (referrer or patient) it very much is a quality metric on overall performance of the operation. Once the time and place of the examination has been chosen (assuming the right test has been chosen in the first place) then the correct, precise protocol should be selected for the indication at hand. For instance, imaging unnecessary body parts only adds to additional radiation (CT) or scan time (MRI). Use of IV contrast may be appropriate for malignant disease but inappropriate for other clinical indications. Increasingly precision protocoling needs to be tailored to the individual patient, their condition, and the question being asked by the referring physician. Protocol appropriateness is a particular problem for many departments as most departments use their own idiosyncratic protocols and there is pervasive variation across institutions and even within departments (some radiologists prefer different protocols for the same clinical indication compared to their colleagues). It is well known, for instance, that radiation dose for the same indication can vary by as much as tenfold depending on the institution, frequently three- to fourfold. Almost no two academic medical centers have similar protocols for the same indications, some with 20, 30, or 45 min MRI protocols for the same indication, for instance.

Similarly modality operations vary considerably from one institution to another. What is seen as an efficient use of assets in one organization is seen as inefficient in another. For instance, one organization may view their 8 a.m.–5 p.m. operation as very busy and productive yet another will operate their scanner from 6 a.m. to 11 p.m. Others will operate their scanners with multiple resources to help expedite patients in and out of the scanner while other organizations will use a single technologist to maneuver the patient on and off the CT table, operate the scanner, send the images to PACS, and go to the waiting room to collect the next patient (a markedly inefficient manner with which to operate an expensive asset). These differences reflect the quality of services as an inefficient operation leads to reduced patient access to scanning (prolonging time to diagnosis) and delays once at the imaging suite (an inconvenience to patients). Once images are generated, radiologists will need the comprehensive set of prior images necessary to determine any new or chronic findings. This has become particularly challenging as organizations consolidate (an increasing trend in the USA) whereby images reside on different and disparate PACS systems which are often poorly connected, if at all. This undermines precision reporting and ability to avoid unnecessary additional imaging tests downstream.

Reporting variation is also widespread both between and within institutions. Even within academic medical centers there is considerable variation of imaging interpretation and analysis of the findings. The reasons are numerous, but imaging has become too complex and sophisticated for any single radiologist (even subspecialty radiologists) to be familiar with the appearances of each disease from each modality with a given protocol. Furthermore, the style and language used by radiologists varies markedly. It is not uncommon for radiologists to offer a range of differential diagnoses without any particular weighting to the chance of one diagnosis being more likely than another. Even the terms to infer degree of risk for a disorder vary from one radiologist to another—one radiologist may believe consistent with confers 100% likelihood of disease, another less so (other terms such as likely, suspicious for, concerning for, also have different connotations from one radiologist to another). There is also widespread evidence that the recommendations made by radiologists for additional tests (especially further imaging) vary considerably in both frequency and type. One radiologist’s certainty for a particular imaging finding may be sufficient for them to recommend no further tests; another may believe a confirmatory and clarification test is required. For instance, some radiologists who diagnose a hepatic hemangioma by ultrasound may stop there; others might recommend a confirmatory additional test such as a three-phase contrast enhanced CT, while others may recommend an MRI. Yet others will recommend further tests immediately; others at a later date. Some might leave the recommendation vague with terms such as consider follow-up MRI (or CT) for example or, even worse, use the term clinical correlation required. In short, this variation reduces the radiologist’s ability to deliver an actionable report and further undermines the rest of any quality improvements implemented in workflow upstream (as indicated by the imaging value chain). Needless to say, referrers and increasingly patients (who often have ready access to their reports) are frustrated by ambiguous and vague narratives and the sometimes frivolous use of further imaging recommendations. These unwarranted variations in practices serve to undermine all other quality and safety efforts.

Once a report has been generated, the referring physician (and patient) expects to receive that report (hopefully actionable) as soon as possible, ideally electronically (so as to be available to as many caregivers as needed). Yet communication standards vary widely too, with some radiologists calling referrers immediately on some reports but not on others, some for routine findings, and some only for critical findings. One might imagine that critical finding alerts (those reports which must be delivered to appropriate caregivers immediately to prevent serious patient harm) should be consistent between institutions (given they are potentially lifesaving in the immediate short term). Yet most departments have critical finding report communication protocols that differ (albeit slightly) from one department to another. Certainly national recommendations exist. Nonetheless, widespread and uniform implementation of the national guidelines has not been achieved.

In summary quality and safety should not just be framed, as it frequently is, around events that lead to obvious immediate harm (i.e., contrast reactions or interventional complications)—rather they should be framed around the myriad of unique activities that constitute the overall radiology workflow. Many of these activities may, in themselves, seem trivial as to their contribution to risk and adversity (such as the recommendation for an unwarranted test, or minor variations in MRI protocol design) but in aggregate these variations can lead to considerable costs and even harm. Until radiologists recognize that it is the responsibility of the overall team to evaluate every operational activity to determine if it meets best practice standards, quality and safety efforts will be undermined and sometimes ineffective. It is the role of leadership to frame the issues to their departments, then build the teams to create, deliver, and manage solutions using data-driven management techniques and the necessary tools and resources to perpetually drive towards better practices and outcomes.

The chapters in this book help address the quality and safety agenda in a systematic and logical order around the concept of the imaging value chain. Subsequent chapters begin with imaging appropriateness (and the use of clinical decision support tools to establish adherence to national guidelines). Chapter III will address protocol optimization with the informed use of medical radiation in diagnostic imaging and further discuss guidelines and standards for managing radiation dose. Chapters will then address modality operations and use of clinical guidelines, image interpretation, structured reporting, and decision support tools for radiologist reporting. Report communication standards will be addressed. Measurement tools and appropriate use of data that have practical and meaningful implications for management of departmental quality and safety will be addressed. Furthermore peer learning and peer review strategies will be outlined that encourage the development of the culture of safety. Finally the emerging field of big data and data analytics to manage the quality and safety agenda will highlight the increasing use of IT systems to drive performance in radiology.

© Springer International Publishing AG 2018

Lluís Donoso-Bach and Giles W. L. Boland (eds.)Quality and Safety in ImagingMedical Radiologyhttps://doi.org/10.1007/174_2017_164

Guideline Development

Michael Bettmann¹, ²   and Myriam Hunink³, ⁴, ⁵  

(1)

Wake Forest University School of Medicine, Winston-Salem, NC, USA

(2)

American College of Radiology, Reston, VA, USA

(3)

Erasmus University Medical Center, Rotterdam, The Netherlands

(4)

Harvard T.H. Chan School of Public Health, Boston, MA, USA

(5)

Netherlands Institute for Health Sciences (NIHES), Rotterdam, The Netherlands

Michael Bettmann (Corresponding author)

Email: [email protected]

Myriam Hunink

Email: [email protected]

References

Abstract

The reason to develop and use Clinical Imaging Guidelines (CIG) is, simply, to improve patient care. CIG development and use are based on the principle that imaging use is not always optimal; reasons for this include lack of expertise (with either the clinical concern or the available imaging modality), nonmedical reasons for requesting the study (medicolegal concerns, possible financial gain), lack of available resources, and expediency. The use of CIG potentially helps in all of these areas, by providing guidance based on high-quality literature and supplemented by expert opinion. CIG answer the question: which, if any, imaging study would be most helpful in this specific clinical situation? The answer is based on assessment of the risk–benefit ratio for the patient. Benefits of imaging are often obvious. Risks, however, also exist, and include the effects of radiation (admittedly difficult to quantify), complications due to contrast agents or the technology (e.g., MRI-related accidents) or other medications, and the possible consequences of unexpected incidental findings that may require evaluation and intervention. The cost to society as well as top individuals must be considered.There are clear steps in the development of any clinical guidelines. First, there must be a sound, reproducible, transparent methodology. Then, specific clinical conditions must be defined, including consideration of their incidence, impact (e.g., success of diagnosis, treatment, and outcomes), and cost to the system. There must be sufficient high-quality literature available to justify review and guideline creation. The literature must be comprehensively and systematically reviewed and summarized. The summary and the topic as a whole must be reviewed by a group that includes all relevant stakeholders. Recommendations must be based to as a great an extent as possible on the literature, supplemented by expert opinion. The guideline must be regularly updated. Any potential conflicts of interest must be clearly presented. Overall, the methodology must follow accepted norms and be reproducible and transparent.There are many challenges in developing and using CIG. In addition to accurate representation and synthesis of what is known, these include adaptation to specific groups of patients and medical systems—what is appropriate for a pediatric age group, for example, may not be appropriate for adults. What works in a fee-for-service developed nation with all imaging modalities available may not work in a rural society with limited equipment and expertise. Finally, the primary goal of CIG is to improve patient care, so they have some clear value for educational purposes, of trainees, non-imagers, patients, families, and regulators. Their greatest use lies in incorporation into the process of requesting imaging studies, to guide appropriate use. This includes prevention of overuse and also elimination of under-use. As such CIG are now widely used in the physician order entry component of electronic health records

The development of clinical imaging guidelines is, by consensus, based on several principles: there is sufficient data from high-quality literature on which to base guidelines, the clinical issues addressed are important, and quality of care can be improved by the development and use of guidelines (Eccles et al. 2012; Committee on Standards for Developing Trustworthy Clinical Practice Guidelines 2011; World Health Organization (WHO) 2012; Bettmann et al. 2015a). There are also several terms used for Clinical Imaging Guidelines (CIG): referral guidelines, appropriate use criteria, appropriateness criteria, and justification guidelines. Although there may be subtle differences, all refer to guidance documents that are developed using a widely accepted and well-defined methodology to provide advice on which specific imaging study, if any, is likely to be most useful in a specific clinical setting. Inherent in guidelines is the focus on balancing the possible benefit against the possible risk. The benefits of imaging range from improved care to reassurance, for the healthcare provider and the patient. There are also both real and potential risks, although in general they are more limited for an imaging study than for a medication or a surgical intervention. The real risks include those related to injury from a medication that may be necessary for the imaging study (e.g., a contrast agent or a sedative) and discovery of unexpected findings that may lead to further investigation or even to intervention but no real benefit to the patient. Findings such as an incidental thyroid nodule (Hoang and Nguyen 2017) or ovarian cyst or a benign liver lesion fall into this category. While significant incidental findings occur, nonsignificant but concerning ones are more frequent (Hoang and Nguyen 2017). Radiation is another risk (Tran et al. 2017; Mathews et al. 2013; Hendee and O’Connor 2012). Although the precise risk of a single imaging exam is essentially impossible to quantitate, and the risk of diagnostic level radiation continues to be debated, it is clear that there are at the very least potential negative consequences of radiation; the concerns are greater in younger patients, due to the latency of these potential adverse effects. Finally, the cost of the imaging exam is an important variable. Depending on the nature of the healthcare system, this may not be a concern to the individual undergoing the imaging exam, but it is always a concern to the system as a whole. If the likelihood is very