1 Introduction to Noninterpretive Skills

Quality: It’s All about Quality Improvement

The NIS primarily concerns itself with core quality and safety.

Quality control (QC), quality assurance (QA), and quality improvement (QI) sound kind of similar, but they are not the same thing. The most up-to-date process is QI. This is the only one of the three that results in an improved system.

Tip: Remember that QI is the current and best iteration because it is the only one with “improvement” in the title.

Quality Control

QC defines the acceptable parameters. For example, the QC of a lumbar puncture includes that “sterile technique must be strictly followed.” The QC won’t tell you the best way to confirm sterile technique, and if infection does occur, the QC won’t give blame or provide fixes. The QC is simply “the rules.”

Quality Assurance

QA is better than QC (although still not as progressive as QI) and is a method of monitoring feedback. For example, telling one of your residents, “Peter, you overcall cardiomegaly on chest X-rays all the time,” is a form of (fairly blunt) quality assurance. Peter may take some solace in knowing that your comment nicely exemplifies key points of QA: reactive, usually retrospective, involves policing (and sometimes punishing) and determines who is at fault. QA is a stable process and, by itself, doesn’t necessarily bring about change.

Quality Improvement

QI is a method to actually change the problematic process. For example, telling Peter the resident, “For the next 10 chest X-ray cases, you need to measure and confirm an enlarged cardiothoracic ratio before concluding that the heart is enlarged,” is a simple form of QI. This method is the best, since it results in change. QI is both retrospective and prospective, avoids attributing blame, creates systems that prevent errors from happening, and is a more recent phenomenon in healthcare.²

A Few Organizations and a Few Acronyms

The Institute of Medicine (IOM) is a nonprofit organization that provides independent and objective analysis and advice to inform public policy decisions related to science, technology, and medicine.³ The IOM has defined six quality aims for medical care, and you have to know this list.

The six quality aims of the IOM are as follows:

1. Safe
   
2. Timely
   
3. Equitable
   
4. Effective
   
5. Efficient
   
6. Patient centered

The acronym STEEEP is provided by the institute to remember these quality aims.

Tip: If you find it tough to remember what this acronym stands for, especially with three nonspecific “Es” in a row, try this mnemonic (Fig. 1.1).

Fig. 1.1 Mnemonic device for the six quality aims of the IOM. Effective, efficient, safe, patient-centered, timely, and equitable.

The Accreditation Council for Graduate Medical Education (ACGME) and the 24 American Board of Medical Specialties (ABMS) agree that a physician must demonstrate the following six core competencies for Graduate Medical Education (GME) and Maintenance of Certification (MOC). Here is the second list you must know.⁴ (In fact, all board-certified physicians in the United States must know this list.)

The six core competencies of MOC are:

1. Patient care
   
2. Medical knowledge
   
3. Interpersonal and communication skills
   
4. Professionalism
   
5. Systems-based practice
   
6. Practice-based learning and improvement

The ABMS provides the mnemonic “Knowledge CLiPS.”

Tip: Try this alternative mnemonic (Fig. 1.2).

1. Patient care: Provide compassionate, appropriate, and effective treatment.
   
2. Medical knowledge: Demonstrate knowledge about established and evolving healthcare science.
   
3. Interpersonal and communication skills: Demonstrate skills that result in effective information exchange with patients, families, and professional associates, and use effective listening skills to work as both a team member and at times a leader.
   
4. Professionalism: Demonstrate a commitment to carry out professional responsibility, ethics, and sensitivity to diverse populations.
   
5. Systems based practice: Demonstrate awareness of the larger context and systems of healthcare, and optimally utilize those systems to improve care.
   
6. Practice-based learning and improvement: Demonstrate the ability to investigate and evaluate patient care practices, assess scientific evidence, and improve the practice of medicine.

Best practice is a method that consistently shows results superior to those achieved with other means, and that method is used as a benchmark.5 “Best practice” has recently become something of a buzzword in business. Best practice examples for employee satisfaction might include creating a culture of engagement, promoting health and well-being, and reducing employee stress.

Tip: A best practice continually evolves.

Dashboards

Dashboards are visual displays of the most important information consolidated into a single screen.⁶ We might think of this as radiology specific (for example, radiology case list, relative value units [RVUs] completed today, and permitted turnaround time per study), but dashboards have much broader utility for any business. A good dashboard can save a company time and money, if the dashboard is exactly tailored to the required industry specs.

Tip: All quality dashboards have immediacy, intuitiveness, and simplicity.

Benchmarking

Benchmarking is the measurement of the quality of an organization’s policies and strategies and their comparison with standard measurements. One goal of a benchmark is to determine where improvements are needed. Another goal is to analyze how other organizations achieve their success.⁷

Tip: In radiology, turnaround time, critical findings compliance, and peer review expectations are examples of benchmarks.

Value

Driven by economics, “value” has grown into a popular medical industry term.⁸ As it relates to healthcare, value is defined as the following:

Value = (service × quality)/cost.

A good value from a patient’s perspective might include interacting with polite radiology staff and receiving quality image interpretation at a reasonable price.

Radiology example

Cash-pay whole-body screening MRI is probably not a good value. The yield is generally low and the cost is generally high.

Real-life example

A $200,000+ European sports car may be of very high quality, but purchasing one is probably not a good value. Top-notch cars like these generally have a low “benefits-to-cost” ratio. Put another way: value = outcomes/cost. For our purposes in radiology, value is directly proportional to optimal appropriateness, good outcomes, improved efficiency, and cost reduction.

It is worth noting that the formerly voluntary physician quality reporting system (PQRS) is now an actual metric used to determine pay.

Tip: Centers for Medicare & Medicaid Services (CMS) is starting to pay for value rather than simply for volume.

Key Performance Indicators

Key performance indicators (KPIs) are business metrics used to evaluate factors that are vital to the success of an organization.⁹ These include both financial and nonfinancial measures integral to defining an organization’s mission and strategy.

KPIs help us measure how well companies, business units, projects, and individuals are performing compared to their strategic goals and objectives. Well-designed KPIs provide the vital navigation instruments that clearly demonstrate current levels of performance.

Radiology example

KPIs might include customer satisfaction, average call wait time, and call abandonment.

Tip: You can’t manage what you can’t measure.

Business Intelligence in Practice

Your office manager informs you that customers complain about their MRI experiences. She says that patients feel like they aren’t getting a good value. They like the quality of the reads, but think the staff are rude, scheduling takes too long, and the co-pays are too high. These are some of the KPIs.

This is the standard cycle for testing changes in QI (and is basically the scientific method). Plan the test or objective, do the test, study the results, and then act (either adopt the change, modify, and try again, or abandon the change). The PDSA cycle, also known as the improvement cycle, is a powerful tool for learning.10

Tip: The ABR advocates using the PDSA cycle for their practice QI (PQI) projects.

You can use PDSA cycles to test an idea by temporarily invoking a change and assessing its impact.

Example: Your patients complain that the information sheet you require them to fill out is too long. You then review the sheet and identify possible unnecessary questions (Plan). Then, you make some modifications to shorten the sheet (Do). Then, you test the new sheet out on a small number of patients. You also make sure that you and the other interpreting radiologists are still satisfied (Study). Then, if the new sheets are satisfactory, you incorporate them into a larger segment of the practice. Finally, if the results of the larger study remain satisfactory, you implement the new sheets practice-wide (Act).

Tip: Start small with the PDSA cycle. Initially, use a limited sample population to make sure your solution works. In the long run, this saves time, money, and risk.

Lean

Lean is the “relentless elimination of waste.” Lean can also be thought of as a method of “making obvious what adds value by reducing everything else.” The Lean principles of process improvement were first described in the manufacturing industry with Toyota automobiles, but translate very well into operations in general. Lean focuses on a smooth workflow and relies on engaging the whole workforce.¹¹ This technique lends itself nicely to radiology, where many people work as a team to produce the final product. Core principles of Lean include eliminating waste, engaging and simplifying, and respecting all people (i.e., customers, employees, and suppliers).

Tip: With Lean, if you respect all people, then you don’t need to remember the “important ones.”

Some specific elements of Lean are discussed below.

“Pull”

This means do the work only when the next step is ready. There should be no work in progress inventory.

Example: Reading a follow-up positron emission tomography/computed tomography (PET/CT) scan without comparison on a patient after his third round of chemotherapy is not optimal. Clearly, there is a comparison around somewhere. Here, the next step is not ready and the comparison should have already been obtained.

“Just in time”

This means to deliver materials only when needed.

Example: Most radiology departments staff the number of radiologists per shift based on historic and usually predictable study volumes. If your CT tech instead batches 15 trauma CT scans, this creates both turnaround time and physician stress problems. Here, the CT tech batched studies rather than delivering them on a regular schedule.

“Kanban”

Kanban means maintaining inventory levels. Processes must have a signal to indicate when supplies need replenishing.

Example: You are scrubbed in the middle of a lumbar puncture on an obese patient and suddenly decide that you need a longer spinal needle. But when you ask for one, the tech explains that there aren’t any and that someone probably used the last one. Here, this preventable problem needed a “Kanban” mechanism to signal low supply.

“Standard work”

Sigma is the Greek letter often used in statistics and refers to the normal distribution of mathematical outcomes.13

“Six Sigma” targets an extremely low defect rate of 3.4 per million opportunities, which represents six standard deviations (SDs) from the population average. In percentages, this means that 99.99966% of the work is done correctly.

As a reminder, one SD equals 68%, two SDs equal 95%, and three SDs equal 99.7% of the data.

Six Sigma is an extremely lofty goal. The real point is to strive for near perfection.

Tip: Like all QI, Six Sigma work is never completed, because a system is never truly “perfected.”

Define, Measure, Analyze, Improve, and Control (DMAIC) Methodology

This data-driven quality strategy is an integral part of Six Sigma, but can also be a stand-alone process.¹⁴

Tip: Lean and Six Sigma share similar methodologies and tools, but there are differences. Lean management is focused on eliminating waste, while Six Sigma management is focused on eliminating defects.

The Eight Wastes in Lean

The standard acronym for these wastes is TIM WOODS.¹²

Value Added or Non-Value Added

Every activity conducted in a business is either value added (VA) or non-value added (NVA). VA activities are those that physically transform a product or service in the eyes of a customer. NVA activities are those that do not. Although some of these may be essential “behind the scenes” activities, they are still considered NVA and encompass waste. The following section discusses examples of NVA activities (waste).

Transport

Movement of product should be controlled by time and distance.

Example: Waiting for hospital transport to bring the patient to radiology is often the rate-limiting step.

Another example: Placing an ultrasound room physically distant from the radiology reading room wastes both sonographer and patient time.

Inventory

Producing inventory that does not immediately move to the next process is NVA.

Example: An interventional radiology department with inaccurate supply need estimations will waste money on expired products.

Motion

Excessive movement of operator and equipment is waste.

Example: A suboptimal computer interface requiring extra clicks to load your cases is NVA.

Another example: Driving a portable X-ray machine to patients’ homes may be prohibitively slow.

Waiting

Waiting should be minimized.

Example: Your system requires the ultrasound tech to enter information on a shared computer after the scan, but before presenting the case to the radiologist. The ultrasound tech may need to wait until that computer is free, and this NVA activity costs the company money.

Overproduction

Product should not exceed demand.

Example: Opening up a new but poorly conceived office that results in few referrals and scans is an example of overproduction.

Overprocessing

This waste results from poor tool or product design. If you don’t know what your customer wants, how can you possibly create the most desired product?

Example: The referring foot and ankle orthopedic surgeon only wants a “limited bone scan of the feet and ankles.” He specifically does not want a standard whole-body bone scan (because he doesn’t want to deal with the incidental findings often picked up on a whole-body scan). Performing the standard whole-body scan on his patients takes extra time and upsets your customer.

Defects

Errors are waste. It usually takes less time to get it right the first time than it does to correct the errors.

Example: Making a sloppy error in a dictation as the result of a hurried effort may require dealing with time-consuming, irate phone calls from referring physicians and ultimately lead to a loss of confidence and business from referring services.

There are seven basic tools of QI. These are a fixed set of graphical techniques identified as the most helpful to troubleshoot issues related to quality. These are termed “basic” because they are intended for people with little formal training to find useful.15

Before discussing the seven methodologies of QI, let’s first examine “the five whys.” Most problems have more than one cause and, to really solve the problem, you need to find that underlying cause. You can often identify the source of the problem using the five whys. This begins with the initial question of why something bad happened. This continues with follow-up why questions.

“The Five Whys” in Action

Example: A STAT CT head wasn’t read within the contractually promised 1 hour, and the emergency room (ER) department complains. The simple move is to blame the radiologist on duty, or conclude that the radiology department is understaffed at that time. But maybe there is more to the story.

You ask the radiologist, “Why didn’t you read it?” He responds, “Because the study was incomplete and was missing images.”

You then ask the tech, “Why was the study incomplete?”

The tech at the hospital responds, “Well, I sent it, but I have no way of knowing if all of the images crossed over.”

You then ask the supervising tech, “Why doesn’t the tech have any way of knowing if all of the images crossed over?” The supervising tech responds, “There actually is a way to tell. The tech working tonight is a rotating tech and must not know how to find that information.” Mystery solved. (Here, only three “whys” were needed to get to the root of the problem.)

The supervisor then trains the tech on duty and develops a training plan for subsequent rotating techs.

The seven tools of QI are as follows:

1. Flow chart
   
2. Cause-and-effect diagram

3. Check sheet
   
4. Control chart
   
5. Histogram
   
6. Pareto chart
   
7. Scatter diagram

1. Flow chart: This is the typical algorithm, and is a picture of the separate steps of a process in sequential order (Fig. 1.3).
   

   

   
   

   Fig. 1.3 Example of flowchart. Algorithm depicts steps needed to fix a light bulb.

   
   
2. Cause-and-effect diagram: This is also known as a fishbone or Ishikawa diagram. (Dr. Ishikawa was a professor of engineering at Tokyo University and a pioneer in QI.) The fishbone diagram identifies many possible causes for a problem, helps to structure a brainstorming session, and immediately sorts ideas into useful categories. The five whys are often invoked in a fishbone diagram (Fig. 1.4).
   

   

   
   

   Fig. 1.4 Example of cause-and-effect diagram. Also known as a fishbone, this diagram helps to brainstorm the possible causes of a missed lung nodule.

   
   
3. Check sheet: This is a structured, prepared form for collecting and analyzing data (Fig. 1.5).
   

   

   
   

   Fig. 1.5 Example of check sheet. This structured form assists in collecting and analyzing data to explain interventional radiology delays.

   
   
4. Control chart: This is a graph used to study how a process changes over time. The data are plotted in chronological order. Using historical data, a control chart has a central line to illustrate deviations from the average, an upper line for the upper control limit, and a lower line for the lower control limit. This differentiates common causes of variation from special causes for variation (Fig. 1.6).
   

   

   
   

   Fig. 1.6 Example of control chart. This chart assesses the average relative value unit (RVU) for members of a radiology group. Radiologist 1 falls below the normal control limit.

   
   
5. Histogram: This is the most commonly used graph to show frequency distributions in the context of two variables. This represents the distribution by mean. If the histogram is normal, the graph takes the shape of a bell curve. Any variation from the bell curve is abnormal, and may take an array of shapes, such as right-sided skewed, bimodal, or even dog-food shaped (Fig. 1.7).
   

   

   
   

   Fig. 1.7 Example of histogram. The graph shows the distribution of daily mean RVUs for 40 radiologists.

   
   
6. Pareto diagram: A Pareto diagram is a bar graph. The lengths of the bars represent the frequency or cost (i.e., time or money) of each variable and are arranged with the longest bars on the left and the shortest bars on the right. This graph visually depicts which situations are most significant, permitting appropriate channeling of resources to fix the problem (Fig. 1.8).
   

   

   
   

   Fig. 1.8 Example of Pareto diagram. This graphically depicts the most significant causes of customer complaints. Types of customer complaints: Long wait (40%), unpleasantwaiting room, delayed appointment scheduling, rude staff, outdated office decor, other. In this example, customer complaints could be decreased by 80% if only the first three problems are addressed.

   
   
7. Scatter diagram: A scatter diagram graphs pairs of numerical data, with one variable on each axis, to look for a relationship between them. This can be useful when trying to identify potential root causes of problems (Fig. 1.9).
   

   

   
   

   Fig. 1.9 Example of scatter diagrams. They demonstrate a good correlation between the dose of contrast A and reactions, but a poor correlation between the dose of contrast B and reactions.

In general terms, one of the bigger building blocks of the NIS is safety. Although safety concerns certainly didn’t originate in 1998, that year a landmark article was published (by the Institute of Medicine [IOM]), which revealed a shocking number of deaths resulting from medical-related errors. This article, entitled, “To err is human: building a safer health system,” brought about many positive changes in healthcare.17

The article had two key points. First, many deaths per year (44,000-98,000) were attributable to medical errors. For perspective, more people die in a given year as a result of medical errors than from either motor vehicle accidents (43,000) or breast cancer (42,000).

Second, the cause of these errors usually wasn’t bad people working in good healthcare systems, but rather, good people working in bad healthcare systems.

Tip: Most errors are preventable with operational changes.

Safety versus Error

Safety and error are a little bit different. Safety is defined as “freedom from accidental injury.” Error is defined as “failure of a planned action to be completed as intended,” or “use of the wrong plan to achieve an aim.”

The Hidden Troubles with Errors

In addition to potentially harming the patient, errors cost money and injure the public’s trust in the system.

On a more personal level, errors can cause you (as the radiologist) real problems.

Example: Say that you missed a cancer on a CT scan, and that your miss really did lead to the patient’s death. You will probably feel awful. Harboring shame and guilt, you might try to keep it a secret from friends and colleagues, and that adds stress. Plus, there is a decent chance that you will get sued. You are now the “second victim.” Healthcare professionals pay for their mistakes, in part, with psychological discomfort. This is similar to posttraumatic stress disorder, and this type of guilt and stress has been known to cause significant morbidity, including suicide. In these cases, it is important for others who are aware of the situation to console.18

Four Fundamental Factors Leading to Errors

The IOM article concluded that four fundamental factors typically led to errors. The mnemonic SLLaP was provided for these factors, and this stands for system, licensing, liability, and payment.

Tip: You can also remember SLLaP, as in getting slapped across the face for making an error (Fig. 1.10).

Fig. 1.10 Mnemonic SLLaP (System, Licensing, Liability, and Payment) of the four fundamental factors leading to errors.

System

This refers to the lack of a single system. Medicine is unfortunately often fragmented, or a “nonsystem.”

For example, when a patient shops from doctor to doctor, loss of information is commonplace. Non-system loss of information may also result from causes beyond the patient’s control (e.g., a change in the company insurance plan).

Licensing

Licensing exams have historically neglected to actually focus on medical errors. This problem is currently being rectified (as evidenced by our own NIS exam).

Liability

Legal issues frighten us into concealing our mistakes, so as a group, we usually lose the opportunity to learn from our colleague’s errors.

Payment

This refers to the failure of third-party payers to provide incentives (i.e., financial rewards) for safety.

Safe move: Organizations often name a “safety champion.” That individual is responsible for owning the safety processes and overseeing the safety reporting system.

Errors

Adverse Event

This is an injury caused by medical care and does not necessarily imply error or blame.

Examples: Pneumothorax post thoracentesis, headache following lumbar puncture, and anaphylaxis after intravenous contrast are adverse events.

When an adverse effect occurs, the system must undergo analysis to detect potential spots for improvement.

Preventable Adverse Event

This is an avoidable event in a particular set of circumstances.

Example: An infected central line as a result of poor sterile technique at the time of the line placement is a preventable adverse effect.

James Reason is a human error psychology expert. He notes that human error is a social label that implies that the individual should have acted differently and is therefore responsible. This is the basis of the “shame and blame” culture that has been so common in healthcare. Shame and blame is a roadblock to process improvement.20

If we understand the root causes of these human errors, and there are many, we should be able to improve. Some of these causes include communication failure, lack of effective training, memory lapses, inattention, poorly designed equipment, fatigue, ignorance, and noisy or distracting working conditions.

Example: Reading screening mammograms in a secluded batched format (away from noise and distractions) improves accuracy.

Malpractice

Quiz: You miss a large pneumothorax on a chest X-ray, and this blatant miss is clearly below the standard of care. However, the ER doctor catches it right away, places a chest tube, and cures the patient. Did you commit malpractice?

Answer: No. See below.

Error and malpractice are different. You might make a whopper of an error, but unless the following four criteria are established, that error does not constitute malpractice.¹⁹

The four criteria needed for malpractice include:

1. A doctor-patient relationship existed.
   
2. The doctor was negligent.
   
3. The doctor’s negligence caused the injury.
   
4. The injury led to specific damages.

In addition, from a legal standpoint, the radiologist is generally obligated to meet the “standard of care,” that is, what would the typical radiologist have done in the same situation?

Errors in detection are the most common cause of radiology malpractice suits. Other causes of malpractice suits include errors in interpretation, errors in communication of results, and errors in follow-up testing suggestions.

Radiology Errors

The two broad categories of error in radiology are perceptual and cognitive.¹⁶

Perceptual Error

The most common radiologist error is a perceptual error. This is a missed finding and accounts for 70% of our errors.

Example: A perceptual error is the pulmonary nodule that you see immediately when you are asked to review the case, but for some reason, you didn’t see or dictate it the first time around. These errors are the hardest for us to accept.

Cognitive Error

The next most common radiologist error is a cognitive error. This error is the result of faulty interpretation of information. Secondary errors in the assessment of pretest probability, failure to seriously consider all relevant possibilities, and various forms of bias, may also contribute to cognitive errors.

Example: A cognitive error is the inappropriate further work-up of a clearly lipid-rich adrenal adenoma seen on a noncontrast CT, if you are not familiar with the criteria to definitively make that diagnosis.

Cognitive errors are more amenable to correction than are perceptual errors.

Other types of radiologist error include “treatment errors,” “failure of communication,” and “preventative errors.” A prevention error might include inadequate screening of a patient with marginal renal function prior to administering iodinated contrast.

For any error, the first step in the response is the same, and that is to deal with the immediacy of the situation and the patient. However, the best response after that first step varies, depending upon the type of error.

Active and Sharp, Latent and Blunt

Active and Sharp

Say you read a right knee MRI, but report it as a left knee MRI. You are the final step in the radiology system, and your error is termed an active error. Errors like yours on the front lines occur at the “sharp end” of the process. The sharp end is the last line of defense against an error. This error occurs at points of contact between a human and some aspect of a larger system. Active errors are readily apparent and easy to blame.

Tip: Remember that the sharp end of the process is on the front lines, since a radiologist poking his needle into a patient is clearly working with a sharp end.

Latent and Blunt

Now say you are the MRI tech who delivered the hapless radiologist that same case. You scanned a right knee, but accidentally labeled it as a left knee. You have committed a latent error, and these errors occur at the “blunt end” of the process. The blunt end includes support for frontline workers (e.g., internal operational personnel, policy makers, and leaders). Blunt end staff decisions affect patient care, but less directly than do sharp end staff decisions. Latent errors are the result of decisions made far from the bedside that impact patient care. These decisions may include organizational flaws (e.g., staffing decisions based on fiscal motivations) or equipment errors that make the human and machine interface less than intuitive.21

Example: The referring doctor has trouble with the cumbersome new picture archives and communication system and enters the wrong diagnostic code. The radiologist, uninformed of the correct diagnosis, is not able to specifically address the diagnostic question.

Tip: Latent errors are “accidents waiting to happen.”

Sentinel Event

A sentinel event is an unexpected occurrence involving death, serious injury, or risk thereof. A sentinel event is a big deal and signals the need for an immediate investigation and response. TJC takes these sentinel events very seriously.²²

According to TJC, some of the most frequent examples of sentinel events include the following: delay in treatment, unintended retention of foreign body, operative/ postoperative complication, suicide, fall, medication error, and “wrong patient, wrong site, wrong procedure.” This last event is particularly common with interventional radiology.

Authority Gradient

This refers to the steepness of the command hierarchy. In general, a lower-level employee tends to be intimidated to speak up to or against a higher-level employee, even if that higher-level employee is making a mistake.

High authority gradients exist in most medical operations and impact negatively on safety.

Example: You and 10 other radiologists perform hysterosalpingograms (HSGs) at the same clinic, all with the same X-ray tech. You, for some reason, were not taught to close the plastic speculum at the time of removal from the patient’s vagina. All of the other radiologists correctly close the speculum. The tech hears your patients complain about discomfort, but does not hear the other radiologists’ patients complain. She correctly concludes that you are doing it wrong. However, she is intimidated by you and your position and chooses to ignore the problem. The result is that the procedural technique goes uncorrected and your patients continue to suffer.

Tip: High authority gradients undermine the safety culture as a result of underreported safety and quality problems.

SBAR

SBAR stands for “situation, background, assessment, and recommendation” and is a basic suggested communication framework that deals with an authority gradient. SBAR is largely used to help nurses and technologists frame discussions with doctors.²³

The SBAR technique has become TJC’s stated industry best practice for standardized communication in healthcare. Nurses or technologists are taught to report in concise narrative form.

Example: “Dr. Carducci, this is Howard, the MRI tech from the Eastern clinic.”

Situation: “Here is the situation. We have patient Raj Agrawal for an MRI with gadolinium.”

Background: “He has a creatinine of 1.8 and an estimated glomerular filtration rate (eGFR) of 39. His labs are stable since a month earlier. He is 67 and takes high blood pressure medication.”

Assessment: “My assessment is that although he is at higher than average risk for developing renal problems, we still should be safe to give him the contrast.”

Recommendation: “I recommend that we proceed with the gadolinium, but give extra hydration and decrease the injected dose. Do you agree?”

Disclosing an Error

What should you do when an error occurs? Over the past decade, it has become clear that the best course of action when an error occurs is to be honest with the patient and family.²⁹ Apologize, take the responsibility, and let them know exactly what steps will be taken to prevent similar events from occurring in the future. The very last thing you want is to appear as if you are trying to cover something up.

Tip: When an error occurs, definitely don’t cover it up or alter the medical record.

Evaluate Risk and Adverse Events

Root Cause Analysis

There is a lot more to optimal safety than simply fixing the immediate problem and blaming the directly involved individual.

When an error occurs, it is necessary to play detective and identify the source of the problem. This process is called “root cause analysis” (RCA).

RCA is a tool used to identify active and latent errors. This creates a narrative of the event with a timeline.²⁴ This is also known as “systems analysis.” The first step in RCA is the quick fix to the actual problem.

Example: Say you have a second patient in a two-day interval develop an anaphylactic reaction to the gadolinium injection. The first step is to save the patient. This first step doesn’t solve the long-term problem.

“High-reliability organizations” can and must operate in a high-stress, high-risk environment.25 For example, aircraft carrier operators, firefighters, and the military fall into this category. Lucky us—so does medicine! When a code is called or when you are simply reading a PET/CT, lives are at stake. You can’t afford to let the ultrasound tech asking you a question or both telephones on your desk ringing distract you from successfully completing your top priorities. Those of us working in this type of high-reliability organization must be at the top of our game, even when the stakes are raised and the emotions are flying.

Five high-reliability organization characteristics are as follows:

This is a blame-free environment where individuals are free to report errors or near misses, without fear of reprimand or punishment.26 This begins with the leadership but includes all employees. Communication is key and safety is always the top priority. In a safety culture, people pursue safety on a daily basis and take action when needed. This can sometimes be difficult, since pointing out blame to coworkers is often uncomfortable.

The Agency for Healthcare Research and Quality (AHRQ) has developed a patient safety culture survey to objectively measure the success of an organization’s safety culture.

Just Culture

A safety culture is also complemented by a just culture. A just culture maintains individual accountability by establishing zero tolerance for reckless behavior.²⁷ A just culture distinguishes between human errors/slips, at-risk behavior, and reckless behavior.

Human Error/Slips

This is an inadvertent action, and often the result of current system design and individual behavioral choices.

Example: The portable intensive care unit X-ray just shot didn’t get recorded because the inexperienced X-ray tech pushed the wrong button. This tech shouldn’t be punished as if an intentional or reckless error had been committed. In fact, if the punishment is too severe for the “crime,” the tech may feel the undeserved pangs of the “second victim syndrome.” Human errors and slips should be managed through changes in design, procedures, and training. Action: console.

At-risk Behavior/Taking Shortcuts

This is an action where the person committing the error believes that the risk is justified. At-risk behavior is choosing to do something in a way that unintentionally imposes a chance for harm to occur.

Radiology example: Choosing to read mammograms on a non-Mammography Quality Standards Act and Program (MQSA)-approved monitor because all of the approved monitors are occupied is at-risk behavior.

Real life example: Driving above the speed limit is at-risk behavior.

At-risk behavior should be managed through removal of incentives for the behavior, creating better incentives, and improving situation awareness. Action: coach.

Reckless Behavior/Ignoring Required Safety steps

This is an action that knowingly puts someone in harm’s way. The risk is identified but ignored.

Example: You are a teleradiologist and receive payment per case read. The only cases on the list are PETCTs, for which you have no training. You know you should not read these cancer patients’ scans, but since you want to make money, you slug through them anyway. This is a conscious disregard for a substantial and unjustifiable risk. This should be managed through punitive action. Action: punish.

Tip: Console, coach, or punish based on the crime, not the outcome. Even if nothing bad resulted, reckless behavior deserves consequences.

Human Factors Engineering

Is it really safer to make a hands-free phone call while driving than to make a call with a handheld device? Psychologists in the field of human factors engineering analyze this type of question. These researchers consider human strengths and limitations when designing interactive systems and tools for people.

The ACR has developed practice guideline standards for the communication of diagnostic imaging findings.28 Consider the following questions.

• What should you do if you can’t reach the referring doctor, but there is a critical finding? For example, what if an outpatient brain MRI for headache shows an unexpected intracranial mass, and you conclude that the patient shouldn’t be driving?
  
• Should curbside consults be documented?
  
• Are all of the following satisfactory methods of nonroutine communication: text, instant message, email, and voice message?
  
• Can your tech give the referring service a nonroutine communication on your behalf?
  
• Is the ACR guideline for the communication of diagnostic imaging findings a legally binding document?

Answers to all of these, along with key points of the ACR document, are discussed below.

The ACR Guidelines

This ACR practice parameter is not a legally binding document. But if you deviate significantly, you should probably document why.

An effective method of communication should promote patient care, support the ordering physician/healthcare provider, minimize the risk of communication errors, and be timely.

The Other Guys

There is a reciprocal duty of information exchange, which includes the referring service. Communication is not simply a one-way street.

An imaging request should contain relevant clinical information, a working diagnosis, pertinent clinical signs and symptoms, and ideally a specific question to be answered. All of this is up to them, and this may be out of your hands.

Your Report

First, a physician (not a tech, physician assistant, or registered physician assistant) must generate the report. The report should include patient demographics and relevant clinical information. Any significant patient reaction should be documented, follow-up or additional diagnostic studies should be suggested (when appropriate), and, unless very brief, the report should contain an “impression.” The final report should be proofread, and abbreviations and acronyms limited, to avoid ambiguity. The report is a definitive document, in compliance with the appropriate regulations, transmitted to the ordering provider, and archived according to state and federal regulations.

Preliminary Reports

These may be either written or verbal and should become a part of the permanent record. Significant variation between preliminary and final reports should be promptly reported and documented.

Standard Communication

The interpreting physician should expedite delivery of the report in reasonable timeliness. This is usually delivered to the ordering physician and/or designees.

Informal Communication

We all encounter these sometimes challenging situations.

Example: Say the oncologist knows and trusts you and wants you to review a CT scan that your colleague read. This patient has a history of lung cancer, in remission, with a reported new renal mass. The oncologist wants to know if the mass is new, and if it is more likely metastatic or a separate new primary. To complicate matters, the only comparison he gives you is an outside report that did not describe a renal mass. This form of curbside consult, often in suboptimal conditions, carries inherent risk.

Tip: The key recommendation for informal interpretations is twofold: radiologists should document these interpretations and radiology groups should develop a formalized system for reporting outside studies.

Nonroutine Communication

Poor doctor-patient communication. The follow-up recommendations were not clearly conveyed during the ER doctor and patient discussion. The patient claims the ER did not specifically tell her to get a contrast study or MRI, if the symptoms persisted. The ER service usually is very careful to provide and document follow-up recommendations, but perhaps because this patient was somewhat difficult to deal with, either the conversation or at least the documentation was neglected.

These Swiss cheese-type error layers unfortunately coordinated to permit the delay in diagnosis (Fig. 1.11).

Fig. 1.11 Swiss-cheese-type error layers. Compound buildup of small errors leading to a large complication.

Key Safety Goals

Key patient safety goals, per TJC.³⁰

Use at Least Two Patient Identifiers Before Any Procedure

Satisfactory identifiers include the patient’s name, date of birth, phone number, patient ID number, government-issued photo ID, and last four of the social security number. The patient’s location or room number cannot be used.

Report Critical Results of Tests on a Timely Basis

Label all medications. In radiology, this includes medicines in syringes. Do this in the area where medicines and supplies are set up (Fig. 1.12).

Fig. 1.12 Label all medications, including medicine in syringes.

Communicate Accurate Patient Information

Take extra care with patients who use blood thinners. Record and pass along correct information about a patient’s medicine.

Prevent Infection

Use the hand-cleaning guidelines from the Center for Disease Control and Prevention (CDC) or the World Health Organization (WHO).

Prevent central line-associated bloodstream infections.

The Procedural Radiologist

The procedural radiologist, a qualified assistant (nurse practitioner or physician assistant), or the referring provider, should perform a specific patient assessment. The assessment should include the history and physical, risk assessment for sedation (if needed), and relevant preprocedure lab tests or other diagnostic tests.

Consent

Informed consent is required for all invasive and some noninvasive medical procedures.³¹ Patients have the right to be informed about the procedures and may request to speak with the radiologist, even when local policy doesn’t require the radiologist to actually obtain the informed consent.

Who Can Obtain the Consent?

This can be obtained by the healthcare provider performing the procedure or by another qualified person assisting the provider. Your tech or aid can obtain the consent. Note that the physician or other provider performing the procedure remains ultimately responsible for answering patient questions/concerns.

Who Can Give the Consent?

The patient should give consent (unless mentally incapable or has not reached the locally recognized age of majority). In those instances, or if the patient is otherwise not able to give consent, permission may alternatively be given by a relative, guardian/legal representative, domestic partner, or healthcare provider who knows the patient.

The three keys (Fig. 1.13)

1. Confirm correct surgery on correct patient at correct body part. Attempt to involve the patient in this confirmation. Additionally, identify the items needed for the procedure. Use a standard list to verify availability of the needed items, and then match these items to the patient.
   
2. Mark the procedure site. Use an unambiguous mark, which is preferably standardized throughout the organization. Only the person performing the exam, or a specific designee, should make the mark. The mark should be at or near the correct site and remain visible after prep and draping. The use of adhesive markers is not permitted. On occasion, a mark is not possible (i.e., on a mucosal surface), or advised (i.e., on the skin of an infant, who may be at risk for permanent stain/tattoo). In those cases, adhere to local policies.
   
3. Perform a “time-out.” Time-out procedures are mandatory for all invasive procedures and for some noninvasive procedures. A time-out should be standardized and documented, and should occur just prior to the procedure. A designated team member should start the time-out, and all members of the team must share involvement and ultimately communicate and agree. The patient should be involved in the process, if he or she is able. If multiple procedures are planned for a given patient, multiple time-outs are required.

Fig. 1.13 Keys to universal preprocedural verification.

Medication reconciliation

This is the process of reviewing the patient’s complete medication regimen at a point of transition and comparing it with the regimen being considered for the new setting of care. This should be done on admission, transfer, and discharge.

Six “rights”

Any time medication is administered, six elements must occur. The right medicine should be given to the right patient, via the right route, with the right dose, at the right time, and with the right documentation. This alliterative phrase constitutes the six “rights” of medication administration.

Simulation-Based Training

Simulation is a learning technique that replaces (or amplifies) real-life experiences with guided ones that replicate substantial aspects of the real world. This often takes place in an “immersive” fashion. In medicine, simulation-based learning can help develop healthcare professionals’ skills, while protecting patients from unnecessary risks.³²

Tip: “I hear and forget. I see and remember. I do and I understand.”—Confucius (551-479 BC)

The ACR has developed appropriateness criteria and decision support. The purpose is to assist physicians in making good imaging decisions for various patient clinical situations.33

These criteria rely on the best evidence available, employ the input of radiologists and other medical specialists, and are posted on the National Guidelines Clearinghouse (NGC), a subset of the AHRQ. Topics include diagnostic radiology, interventional radiology, and radiation oncology.

Ten ACR appropriateness imaging criteria topics include breast, cardiac, gastrointestinal, musculoskeletal, neuro, pediatrics, thoracic, urologic, vascular, and women’s imaging. Expert panels comprised of radiologists and clinical specialists outside of radiology maintain the topics, each of which contains various clinical conditions.

Each variant is scored on a 1-9 scale (1 = low, 9 = high).

• 1-3: test usually not appropriate.
  
• 4-6: test may be appropriate.
  
• 7-9: test usually appropriate.

Additionally, relative radiation risk is included for decision support.

Tip: Decision support use will be mandated by CMS for reimbursement in 2017.

Practice Parameters and Technical Standards

The ACR practice parameters and technical standards define principles and technical parameters and describe a range of acceptable approaches for diagnosis and treatment. These are reviewed at least every 5 years and allow for differences in training, experiences, and local conditions. Technical standards are quantitative or measureable and focus on physics and nuclear medicine.

Peer Review

Peer review is required by the ACR to ensure that ongoing quality review processes are in place. There is no punitive aspect, the process is anonymous, and peer review data are immune to legal action or discovery. A variety of biases may hamper the peer review process, and these should be avoided as much as possible.

TJC checks that a peer review process is ongoing but does not review the actual data.

RadPeeR

RadPeeR is a program that many radiology departments use for their peer review. Radiologists typically sample 5 to 10% of each other’s cases. Specialists usually only rate other specialists, and the samples are random.

RadPeeR data is often conveniently linked to ongoing professional practice evaluation (OPPE) and focused provider practice evaluation (FPPE).

RadPeeR data are scored as follows:

1. Concur.
   
2. Discrepancy, but not ordinarily expected to be made. An understandable miss.
   
3. Discrepancy, and diagnosis should be made most of the time.
   
4. Discrepancy, and diagnosis should be made almost every time. A clear misinterpretation of the study.

Additionally, RadPeeR data 2, 3, and 4 are further subdivided into the following:

• Unlikely to be clinically significant.
  
• Likely to be clinically significant.

Tip: QI data are not discoverable in a court of law.

Miscellaneous Organizations Contributing to Quality and Safety

ACR: RadPeeR, modality certification, practice guidelines, and technical standards, BI-RADS (LI-RADS, PI-RADS, etc.), and appropriateness criteria.

American Roentgen Ray Society (ARRS): educational courses, Practice Quality Improvement (PQI) projects.

To minimize the significant ferromagnetic safety risks of the MR scanner magnet, special mandated precautions exist for every MRI department.34 There are four zones and they are all color-coded. Zone 1 (green) is the safest and most removed from the magnet. Zone 4 (red) is actually in the MRI room.

The four MR safety zones are as follows:

• Zone 1: outside (green)
  
• Zone 2: nearby, but no door to MRI suite (blue)
  
• Zone 3: nearby, with door to MRI suite (yellow)
  
• Zone 4: room housing the MRI (red)

Contrast Allergies and Reactions

The two main challenges associated with iodinated contrast administration are contrast-related allergies and contrast-induced nephropathy (CIN).³⁵

Incidence

The incidence of any contrast reaction is 0.2 to 0.7%. The incidence of a severe reaction with low osmolality contrast is 0.01 to 0.02%. The greatest risk factor for a reaction is a history of a prior contrast reaction. This increases the risk of subsequent reaction fivefold. However, even in a patient with a history of prior reaction, and without the addition of a pretreatment regimen, the risk of subsequent reaction is still less than 10%.

Shellfish, Asthma, and Atopic Patients

Atopic patients have a twofold to threefold increased risk of reaction. Asthmatic patients may also have a slightly increased risk. Shellfish allergy patients have an equivocal increased risk of contrast reaction.

Premedication

Prednisone: 50 mg PO 13 hours, 7 hours, and 1 hour prior. Diphenhydramine (Benadryl): 25 to 50 mg intravenous (IV)/ intramuscular (IM) or PO 1 hour prior.

Tip: Note that IV steroids are not effective when administered less than 4 to 6 hours prior to contrast.

What can premedication accomplish?

Premedication is proven to reduce the incidence of minor reactions. Premedication is not proven to protect against severe life-threatening injuries, although the rarity of severe reactions makes it difficult to study this question. For practical purposes, if a patient has a history of a severe reaction (and if after careful risk-benefit assessment, the decision is made to administer contrast), a premedication regimen is recommended.

Breakthrough reactions

Some patients will exhibit breakthrough reactions despite premedication. These reactions are most often similar to the index reaction. If these patients are subsequently premedicated, the large majority won’t have another breakthrough reaction. Patients with a mild index reaction have an extremely low risk of subsequently developing a severe breakthrough reaction. Severe allergies to any other substance are associated with a somewhat higher risk of developing a moderate or severe breakthrough reaction.

Tip: Anaphylactic reactions are unpredictable and life-threatening.

Treating the Radiology Patient with a Contrast Reaction

Mild reaction

If a patient develops urticaria from contrast, no specific treatment is usually indicated. (Discontinue the injection if not already completed.) Benadryl improves the urticaria but also causes mild sedation.

Tip: Any patient with a mild reaction should be observed for at least 20 to 30 minutes, since mild reactions may evolve into more severe reactions.

More-Serious Reactions (Fig. 1.14)

Bronchospasm

Fig. 1.14 Treating patients with contrast reactions.

Give oxygen 6 to 10 L/min via mask. Monitor electrocardiogram (EKG), oxygen saturation, and blood pressure. Give β-agonist inhaler (i.e., albuterol 2-3 puffs, when necessary). If there is no response, give epinephrine.

Give epinephrine subcutaneous (SC) or intramuscular (IM) (1:1,000) 0.1 to 0.3 mL (0.1-0.3 mg) or, especially if hypotension is evident, epinephrine 1:10,000 slowly intravenous (IV) 1-3 mL. Repeat as needed up to a maximum of 1 mg.

Hypotension with tachycardia

Elevate legs 60 degrees or more (or Trendelenburg, which is 15-30 degrees of legs higher than the head). Give oxygen 6 to 10 L/min via mask. Monitor EKG, oxygen saturation, and blood pressure. Bolus rapid IV large volume of lactated Ringer’s or normal saline. If the response is poor, give epinephrine 1:10,000 slowly IV 1 mL (0.1 mg). Repeat as needed up to a maximum of 1 mg.

Hypotension with bradycardia (vagal)

Same initial steps as above, but instead of epinephrine, if poorly responsive, give atropine 0.6 to 1 mg IV slowly. Repeat up to a total dose of 0.4 mg/kg (2-3 mg) in an adult.

Physiologic but Undesirable Effects

These include arrhythmias, depressed myocardial contractility, cardiogenic pulmonary edema, functional hypocalcemia, and even seizures. Contrast problems are often dose and concentration dependent. It is reasonable to consider slightly lowering the dose in a higher-risk patient.

Contrast Nephropathy

This is defined as the impairment of renal function attributable to recent contrast administration. There is some debate on the standard criteria, but many make this diagnosis when there is a 25% increase in serum creatinine from baseline, or a 0.5 mg/dL increase in absolute value, within 48 to 72 hours of intravenous contrast administration. Creatinine levels typically peak within 2 to 5 days and revert to normal within 7 to 10 days. The development of permanent renal dysfunction is rare, but there does appear to be a (complex) relationship between CIN, additional morbidity, and even mortality.

Risk Factors

Age, chronic kidney disease (CKD), diabetes, hypertension (HTN), metabolic syndrome, anemia, multiple myeloma, hypoalbuminemia, renal transplant, hypovolemia, and decreased left ventricular ejection fraction (LVEF) are all risk factors. Low osmolality contrast is less nephrotoxic than high osmolality contrast. Cardiac procedures result in a much higher incidence of CIN than do diagnostic radiology procedures.

Prevention

Giving IV fluids before contrast administration is the most proven method of reducing risk. One suggestion is to begin hydration 6 hours prior to contrast and continue for 6-24 hours afterward. Congestive heart failure and other fluid-intolerant patients pose a special challenge.

A variety of other pretreatment regimens, including mannitol, lasix, theophylline, endothelin-1, and fenoldopam, are less proven, and use of these agents for the prevention of CIN is not recommended.

Using a lower dose of contrast may be helpful. Most authors agree to keep the contrast dose less than 100 mL, since studies have shown that the need for subsequent dialysis when contrast is kept to that dose or lower is very rare.

For a patient with a moderate to severe risk of CIN, the estimated glomerular filtration rate (eGFR) should be estimated precontrast and then remeasured 24 to 48 hours postcontrast. If CIN develops, hydration is the cornerstone of therapy.

Tip: The best treatment for CIN is prevention. Screen appropriately, use the lowest dose of contrast possible, and ensure IV hydration.

Quiz: Is taking metformin a risk factor for developing CIN?

Answer: No, but a patient with renal failure while taking metformin is at risk for developing lactic acidosis.

Contrast Warming

We have come a long way with radiation safety since Mrs. Roentgen first put her hand in front of a cathode ray tube. There are now a number of thoughtful radiation safety programs.36

Radiation Programs

Image Gently

This was initially produced by the Society of Pediatric Radiology, with the goal of raising awareness and minimizing radiation doses in pediatrics. The first focus was on children and their CT scans.

Step Lightly

This was initially introduced by pediatric interventional radiologists, who provided educational resources to further the goal of decreasing fluoroscopic doses. There are three phases with this program: Phase 1 educates the radiologists, technologists, and physicists. Phase 2 educates the referring services. Phase 3 educates the parents and public.

Image Wisely

This is the adult counterpart of Image Gently and was formed by the RSNA and the ACR. The focus is twofold: one, optimizing protocols and appropriate use of imaging, and two, communicating with patients, the public, and the medical community. The primary focus thus far has been with CT scans.

Patients are now able to track their exams and cumulative radiation doses with individual “patient medical imaging records.”

Considerations When Explaining Radiation Risks to Patients

First of all, it is important to ensure that the exam is appropriate. It is also important to place context in the patient’s diagnosis. Discussion of risk of the exam should also include the risk of not performing the exam. It is often a good idea to compare radiation risk to other radiation risk exposures (e.g., the radiation exposure for traveling in an airplane at 30,000 feet compared to the radiation exposure for the radiology test in question). Try to explain to the patient that risks are discussed in terms of populations and that individual risk, although obviously desired, is very difficult to determine.

CT Dose Index

This is a standardized measure of radiation dose output of a given CT scanner. This permits comparison with the radiation output from other CT scanners. Note that the CT dose index is a dose estimate and does not give the actual dose received by the patient. (The dose received depends on multiple factors including the CT scanner, technical parameters, scan protocol, number of phases, pitch, and tissue composition, size, and frequency. Medical physicists can calculate an estimated dose based on parameters of the study and specific patient factors.)

There are two main parameters calculated and reported by CT scanners:

1. CTDI vol: Volume-based CT dose index (units: mGy).
   
2. DLP: Dose-length product (units: mGy-cm). Note that DLP = CTDI vol × scan length (for helical scanning).

American College of Radiology: Dose Index Registry

Data of CTDI vol and DLP for CT scans performed are tracked and compared to those of other institutions. Facilities may use this data to track their own performance and adjust protocols.

The ACR has provided several reference CTDl vol values. These values let practices know whether their numbers are at least “in the ballpark”:

• 75 mGy for a head CT
  
• 25 mGy for an adult abdomen CT
  
• 20 mGy for a pediatric abdomen CT

ALARA

The proportion of people who have the disease and test positive for it.37

A test with a high sensitivity is most useful for ruling out disease.

You can remember this by SNOUT: Use the test of SeNsitivity to rule OUT disease (Fig. 1.15).

Fig. 1.15 Sensitivity testing. Use the test of SeNsitivity to rule OUT disease (SNOUT).

For example, a chest CT has a very high sensitivity for the detection of a pulmonary nodule. If no nodule is detected, it is very unlikely that the patient has a pulmonary nodule.

Specificity

The proportion of people who do not have the disease and test negative for it.

A test with a high specificity is most useful for ruling in the disease.

You can remember this by SPIN: Use the test of SPecificity to rule IN disease (Fig. 1.16).

Fig. 1.16 Specificity testing. Use the test of SPecificity to rule IN disease (SPIN).

For example, a lower extremity Doppler ultrasound has a very high specificity for the detection of a blood clot. If a clot is detected, it is very likely that the patient indeed has a clot.

Positive Predictive Value (PPV)

The proportion of people with positive test results who actually have the disease (TP = true positive; FP = false positive):

PPV = TP/(TP + FP).

Negative Predictive Value (NPV)

There are many potential sources of error in statistical analysis, and a variety of biases contribute to these errors.38

• Selection bias occurs when data for analysis are compiled in such a manner that proper randomization is not achieved.
  
• Spectrum bias, a subtype of selection bias, occurs when the sample is missing important subgroups.
  
• Verification bias occurs when the results of a diagnostic test affect whether the gold standard procedure will be used to verify the test result.
  
• Sampling bias occurs if some members of the population are either more or less likely to be included in the study.
  
• Measurement bias occurs when methods of measurement are dissimilar between different groups of patients.
  
• Review bias, a subtype of measurement bias, occurs when tests are performed or interpreted without proper blinding.
  
• Confounding bias occurs when two factors are associated and the effect of one is distorted or confused by the effect of the other.
  
• Lead-time bias occurs when a new test merely identifies the disease earlier than did the previous test, thus giving the impression that the survival is longer. (An appropriate way to avoid lead-time bias is to compare age-specific mortality rates, rather than survival rates from the time of diagnosis.)
  
• Length-time bias occurs when screening tests falsely appear to improve survival, since they often detect more slowly growing diseases, which tend to have a better prognosis.
  
• Screening bias occurs because patients who volunteer for screening studies tend to be healthier and have better outcomes than nonvolunteers.
  
• Overdiagnosis bias occurs when screening identifies an illness that would not have shown clinical signs before a person’s death from other causes.
  
• Pygmalion effect bias occurs when higher expectations lead to an increase in performance.

The Hawthorne effect was first described in the work-place.39 A series of experiments was performed at the Chicago AT&T Hawthorne factory in the 1920s. Initially, workers were told that they would be exposed to greater light and that their subsequent productivity would be observed. Indeed, their productivity improved—briefly. But the change in productivity wasn’t the result of better lighting. In fact, the experimenters found that any variable change in the workplace, including break schedules, working hours, or self-selection of nearby coworkers (as long as the workers knew they were being observed) resulted in improved productivity. These effects were all temporary. Further, researchers found that social factors (e.g., working as a team, choosing one’s own coworkers, and perceived interest that the manager had in a project) were at least as important as physical improvements in working conditions on improving productivity.

Knowledge of these effects has had great applications in management. For research, the Hawthorne effect is one of the hardest biases to eliminate from an experiment.

Weber Effect

Adverse event reporting tends to increase in the first two years after introduction of a new agent or use for a new indication, peaks at the end of the second year, and then declines.⁴⁰ This pattern was first described in 1984 in the United Kingdom for a single drug and is known as the Weber effect. The Food and Drug Administration and other agencies have made concerted efforts to improve methods and reporting. Today, this effect is probably not as common. The Weber effect explains why in a PDSA cycle the incidence of an adverse event may initially rise.

Now turn to the next chapter for additional new material and to begin the question and answer format.

References

1. CMS.gov. https://www.cms.gov/research-statistics-data-and-sys-tems/statistics-trends-and-reports/nationalhealthexpenddata/nhe-fact-sheet.html. Published 2015. Accessed May 6, 2016

2. HRSA.gov. http://www.hrsa.gov/quality/toolbox/methodology/qualityimprovement/. Published 2015. Accessed May 6, 2016

3. National Academies. http://www.nationalacademies.org/hmd/. Published 2016. Accessed May 6, 2016

4. American Board of Medical Specialties. http://www.abms.org/board-certification/a-trusted-credential/based-on-core-compe-tencies/. Published 2016. Accessed May 6, 2016

5. Techtarget. http://searchsoftwarequality.techtarget.com/definition/best-practice. Published 2016. Accessed May 6, 2016

6. Techtarget.http://searchbusinessanalytics.techtarget.com/definition/business-intelligence-dashboard. Published 2016. Accessed May 6, 2016

7. Ettorchi-Tardy A, Levif M, Michel P. Benchmarking: a method for continuous quality improvement in health. Health Policy 2012;7(4):e101-e119

8. Porter ME. What is value in healthcare? N Engl J Med 2010;363(26):2477-2481

9. Klipfolio. https://www.klipfolio.com/resources/kpi-examples. Published 2015. Accessed May 6, 2016

10. Institute for Healthcare Improvement. http://www.ihi.org/resources/Pages/HowtoImprove/default.aspx. Published 2016. Accessed May 6, 2016

11. Lawal AK, Rotter T, Kinsman L, et al. Lean management in healthcare: definition, concepts, methodology and effects reported (systematic review protocol). Syst Rev 2014;3:103

12. Six Sigma Online. http://www.sixsigmaonline.org/six-sigma-training-certification-information/lean-six-sigma-wastes-and-principles/. Published 2015. Accessed May 6, 2016

13. American Society for Quality. http://asq.org/learn-about-quality/six-sigma/overview/overview.html. Published 2015. Accessed May 6, 2016

14. American Society for Quality. http://asq.org/learn-about-quality/six-sigma/overview/dmaic.html. Published 2015. Accessed May 6, 2016

15. American Society for Quality. http://asq.org/learn-about-quality/seven-basic-quality-tools/overview/overview.html. Published 2015. Accessed May 6, 2016

16. Sabih DE, Sabih A, Sabih Q, Khan AN. Image perception and interpretation of abnormalities; can we believe our eyes? Can we do something about it? Insights Imaging 2011;2(1):47-55

17. The National Academies of Sciences Engineering and Medicine. https://www.nationalacademies.org/hmd/Reports/1999/To-Err-is-Human-Building-A-Safer-Health-System.aspx. Published 2016. Accessed May 6, 2016

18. Grissinger M. Too many abandon the “second victims” of medical errors. P T 2014;39(9):591-592

19. Bal BS. An introduction to medical malpractice in the United States. Clin Orthop Relat Res 2009;467(2):339-347

20. Reason J. Human error: models and management. BMJ 2000;320(7237):768-770

21. Agency for Healthcare Research and Quality. https://psnet.ahrq.gov/primers/primer/21/systems-approach. Published 2015. Accessed May 6, 2016

22. The Joint Commission. http://www.jointcommission.org/senti-nel_event.aspx. Published 2016. Accessed May 6, 2016

23. Safer Healthcare. http://www.saferhealthcare.com/sbar/what-is-sbar/. Published 2015. Accessed May 6, 2016

24. American Society for Quality. http://asq.org/learn-about-quality/root-cause-analysis/overview/overview.html. Published 2015. Accessed May 6, 2016

25. The Joint Commission. http://www.jointcommission.org/assets/1/6/chassin_and_loeb_0913_final.pdf. Published 2016. Accessed May 6, 2016

26. United States Department of Labor. https://www.osha.gov/SLTC/etools/safetyhealth/mod4_factsheets_culture.html. Published 2016. Accessed May 6, 2016

27. Boysen PG II. Just culture: a foundation for balanced accountability and patient safety. Ochsner J 2013;13(3):400-406

28. American College of Radiology. http://www.acr.org/Quali-ty-Safety/Standards-Guidelines. Published 2016. Accessed May 6, 2016

29. Agency for Healthcare Research and Quality. https://psnet.ahrq.gov/primers/primer/2/error-disclosur3. Published 2015. Accessed May 6, 2016

30. The Joint Commission. http://www.jointcommission.org/standards_information/npsgs.aspx. Published 2016. Accessed May 6, 2016

31. The Joint Commission. http://www.jointcommission.org/assets/1/23/Quick_Safety_Issue_Twenty-One_February_2016.pdf. Published 2016. Accessed May 6, 2016

32. Al-Elq AH, Al-Elk. Simulation-based medical teaching and learning. J Family Community Med 2010;17(1):35-40

33. American College of Radiology. http://www.acr.org/Quality-Safety/Appropriateness-Criteria. Published 2016. Accessed May 6, 2016

34. American College of Radiology. http://www.acr.org/quality-safety/radiology-safety/mr-safety. Published 2016. Accessed May 6, 2016

35. American College of Radiology. http://www.acr.org/quali-ty-safety/resources/contrast-manual. Published 2016. Accessed May 6, 2016

36. American College of Radiology. http://www.acr.org/quality-safety/radiology-safety/radiation-safety. Published 2016. Accessed May 6, 2016

37. Parikh R, Mathai A, Parikh S, Chandra Sekhar G, Thomas R. Understanding and using sensitivity, specificity and predictive values. Indian J Ophthalmol 2008;56(1):45-50

38. Sica GT. Bias in research studies. Radiology 2006;238(3):780-789

39. McCambridge J, Witton J, Elbourne DR. Systematic review of the Hawthorne effect: new concepts are needed to study research participation effects. J Clin Epidemiol 2014;67(3):267-277

40. Hoffman KB, Dimbil M, Erdman CB, Tatonetti NP, Overstreet BM. The Weber effect and the United States Food and Drug Administration’s Adverse Event Reporting System (FAERS): analysis of sixty-two drugs approved from 2006 to 2010. Drug Saf 2014;37(4):283-294

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Questions and Answers Doppler Histogram Vessel Density Analysis The Normal Brain Scan Lymphoproliferative Diseases Meningioma—Viewpoint: Fractionated Radiotherapy Accounting in Radiology

Stay updated, free articles. Join our Telegram channel

Join