- Current Issue
- ACP HospitalistWeekly
- Career Connection
- Renew Your Subscription
- RSS Feeds
- Write for ACP Hospitalist
Glutted with guidelines?
Which ones to trust and how they can help you
By Janet Colwell
Imagine you are treating a patient for condition “Y,” and the relevant clinical guideline recommends treating with drug “X.” But the drug is not on your hospital's formulary. Do you look for a suitable alternative, or ask the hospital to try to obtain drug “X”?
The decision, as with many in clinical practice, is up to the physician. Guidelines provide valuable direction, but clinicians ultimately must rely on experience and judgment to determine the next step.
Photo by Thinkstock.
“Guidelines should be viewed as a general roadmap, but that map doesn't necessarily tell you the best path to take,” said Juliana Barr, MD, critical care specialist and associate professor of anesthesia at Stanford University School of Medicine in Palo Alto, Calif. “Some of the aspects of applying guidelines depend on interpreting them in the context of a particular patient or clinical situation or culture of the health care environment you're working in.”
When deciding which guidelines to follow, clinicians should look for strong supporting evidence and clear explanations of how the authors arrived at their recommendations, said Dr. Barr, chair of the American College of Critical Care Medicine task force for the 2013 Clinical Practice Guidelines for the Management of Pain, Agitation and Delirium in Adult Patients in the ICU.
Be skeptical when recommendations are based on expert opinion or weak evidence, she advised. Remember, too, that guidelines reflect the evidence available at the time they were written, said Brian Kavanagh, MB, FRCPC, chair of the department of anesthesia at the University of Toronto and a critical care medicine specialist at the Hospital for Sick Children in Toronto. New research can potentially alter or negate specific recommendations.
“Having consistent care with lack of variability is a good thing if you know what the best care is, but sometimes you don't know the best care,” he said. He cited the example of the 2012 International Guidelines for the Management of Severe Sepsis and Septic Shock, which removed a previous recommendation to maintain tight glucose control, based on new evidence that such control was harmful.
“Guidelines should be viewed as tools, as collections of statements and recommendations with variable amounts of evidence, that tell you how to navigate through the complexities of a clinical management issue,” said Dr. Barr. “Where people often get hung up on guidelines is when they interpret them in absolute terms and see them as black-and-white rules that must be followed.”
Grading the evidence
Clinical guidelines are often held up as the gold standard for evidence-based practice, but experts have raised concerns in recent years that not all are backed by high-quality or conclusive evidence.
In 2011, Archives of Internal Medicine published an analysis of guidelines issued by the Infectious Diseases Society of America (IDSA) finding that more than half of the society's current recommendations were based on expert opinion alone. Similarly, a 2009 analysis in the Journal of the American Medical Association assessed cardiovascular guidelines issued by the American Heart Association/American College of Cardiology and found that most recommendations were based on lower levels of evidence and expert opinion.
Those concerns have helped fuel a movement over the past decade to improve the way guidelines are derived and written, resulting in the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system.
Most recent guidelines use the GRADE system, which classifies recommendations as strong or weak based on the strength of the evidence and patient outcomes. Many high-quality randomized controlled trials (RCTs) that show benefit from a particular intervention merit a strong recommendation. But a recommendation from a robust study may be weakened if there is no clear evidence of patient benefit.
For example, if there is good evidence that a particular drug is associated with a high cancer-survival rate, but the drug also is associated with a high degree of morbidity and complications, it might result in a weak recommendation, Dr. Barr said.
In addition, the GRADE system does not allow for expert opinion (defined as evidence that is based only on clinical experience and not backed by any research studies) to be used as a substitute for high-quality evidence. “In the absence of evidence, or if there is conflicting evidence, no recommendation is made,” Dr. Barr said.
GRADE's succinct summaries of all supporting literature help physicians with the challenge of deciding when and how to apply a recommendation in their own practice, said Ole Vielemeyer, MD, an infectious diseases specialist and assistant professor at Drexel University College of Medicine in Philadelphia. “If guidelines are not well written or do not include all of the existing literature, they may mislead you,” he said.
When to follow, when to diverge
With the push to standardize care, some have suggested that hospitals make it compulsory for clinicians to follow accepted clinical guidelines. However, many experts maintain that guidelines should provide direction, not mandates.
“Making guidelines available for people to refer to is a good idea,” said Dr. Kavanagh, “but making clinicians who actually know what they're doing follow them doesn't make a lot of sense, because in many situations in medicine, you want physicians to be reappraising all the time.”
Focusing too heavily on following guidelines, rather than treating the patient in front of you, can discourage creative thinking, he added.
“Say a hospitalist gets a handover of a patient with a diagnosis of asthma, but the patient actually has heart failure. Instead of being reassessed all the time as to what is wrong with them, they are only reassessed as to whether the guideline is being implemented,” said Dr. Kavanagh. “The clinician needs to say, ‘Do I have the diagnosis right and am I giving the best treatment?’, not just accept the diagnosis.”
Also, guidelines that work for most people may not fit a particular setting. The workings of a large hospital with a highly specialized intensivist team are not the same as a small community hospital, nor are facilities with an open versus a closed ICU model the same. Physicians should be allowed to adjust their practice accordingly, said Dr. Vielemeyer, co-author of the 2009 analysis of IDSA guidelines. “Guidelines are useful and powerful but we have to understand their limitations.”
The danger of following recommendations based solely on expert opinion is apparent in the case of hormone replacement therapy (HRT), said Dr. Barr. For decades, postmenopausal women were advised by experts to take HRT—until the Women's Health Initiative provided evidence that HRT didn't confer the expected health benefits and might even be harmful.
“It took decades to debunk that myth,” said Dr. Barr. “And in the meantime there were untold negative health costs borne by women who were blindly being treated with HRT.”
Importance of local control
Once a hospital has decided to adopt a set of guidelines, it's crucial to engage physicians and other front-line clinical staff in the implementation process, said Bruno DiGiovine, MD, MPH, head of pulmonary, critical care, and sleep medicine at Henry Ford Hospital in Detroit.
For example, a hospitalist group might agree to implement protections against central line-associated infections, such as full barrier precautions, hand washing and chlorhexidine skin antisepsis, he said. “But then your group decides how exactly to do it—how will you use a checklist, what rounding tools might you use, who's going to look at the checklist, etc.—to make it all work in your local environment.”
Physicians also should be involved in measuring processes and outcomes associated with guidelines, said Dr. Barr. One reason that clinicians may not follow guidelines is that they believe they are already following best practices, she said. Only by measuring processes can hospitals know whether guidelines are actually being applied in practice and, if not, find out why.
Measuring outcomes is trickier because the relatively small numbers of patients treated for one condition at one hospital often don't generate enough data to draw valid conclusions, said Judith Jacobi, PharmD, critical care pharmacy specialist in the adult critical care units at Indiana University Health Methodist Hospital in Indianapolis. However, there are some quality measures that can be used to assess whether a particular process or protocol is effective.
For example, you can measure whether blood glucose stayed below a threshold level after implementing insulin therapy guidelines, said Dr. Jacobi, who chaired the Society of Critical Care Medicine's task force on the 2012 Guidelines for Insulin Infusion Therapy in Critical Care. In the case of sedation, you can measure whether compliance with a target sedation level or daily drug holidays resulted in the desired outcome of avoiding oversedation.
It's important to measure compliance with protocols in order to know whether the processes you've put in place are working, said Dr. Jacobi.
“We put protocols in place to try and achieve what the guidelines have suggested,” she said. “Although we can't measure the same kind of outcomes [as a large RCT], like survival, there are certainly other measures that are typically important in an individual setting.”
In cases where the benefits of a particular treatment are well established by evidence, measuring compliance with the process is sufficient, said Dr. DiGiovine.
“If we know that aspirin is important for patients with myocardial infarction, we just need to measure whether patients are getting one, not the mortality of our patients with MI,” he said. “For other things where we're not sure whether a process leads to an outcome, we may want to measure outcomes as well. In preventing central line infections, for example, we want to measure that we're taking preventive measures, but we would also measure outcomes of whether we're getting central line infections.”
Guidelines as checklists
While guidelines can be flawed, it's important to recognize that they are still powerful tools in everyday practice, experts said.
Guidelines should work as a checklist for physicians, said Dr. Vielemeyer. “They are a reminder of what basic elements of patient care should be present when you are taking care of a certain problem.”
Physicians should use guidelines to help them focus on the most important literature about a particular issue that has been evaluated by top experts in the field, said Dr. Jacobi. They can set a path for organizing your practice more efficiently and provide some processes to measure.
“Using guidelines is not cookie-cutter medicine because we still have to be able to apply them,” said Dr. Jacobi. “We encourage clinicians to look at guidelines not as gospel, not as absolutes, but as an important pathway to care for the majority of their patients.” (ACP recently released a free app of its clinical guidelines on iPhone, iPad and Android.)
Guideline authors acknowledge that the documents they produce won't answer every question, no matter how strong the evidence and analysis, said Dr. Barr. Those gaps in knowledge are opportunities for future research, and physicians need to keep abreast of new developments that may change current thinking.
“You have to understand that from the day guidelines are published they are incomplete and as they move forward and more research is published, it will make some of the recommendations obsolete,” said Dr. Barr. “Physicians have to be flexible and adaptive as new evidence comes along.”
At the same time, don't assume that one new study refuting a guideline's findings is conclusive, she added. “If a set of guidelines says you should do ‘X’ intervention based on 10 studies that all point in that direction, and an eleventh study comes out that refutes that, providers need to be very critical in their analysis of that study and not just jump on the latest thing,” she said.
Most important, physicians must take the time to read and understand what's behind the recommendations they use, said Dr. Kavanagh.
“If you don't read the evidence, you are turning yourself into a technician and you divorce yourself entirely from the underlying medical literature,” he said. “You will feel a lot better about applying the guideline recommendations if you read the underlying evidence and it makes sense to you.”
Janet Colwell is a freelance writer in Miami.
Smart guideline use: One expert's advice
In a plenary talk at the Society of Critical Care Medicine's annual conference, held in Puerto Rico in January, Brian Kavanagh, MB, FRCPC, advised on the intelligent use of guidelines. Dr. Kavanagh is chair of the department of anesthesia at the University of Toronto and a critical care medicine specialist at the Hospital for Sick Children in Toronto. Here are his suggestions:
- Definitely use guidelines if they cover a topic you are unfamiliar with or unsure about. However, if you find you are frequently using the same guidelines, go educate yourself on the clinical issue—read more and learn more, so you can evaluate the guidelines critically and understand the issues better.
- Understand that not all guidelines are created equal. Clinical knowledge comes in many forms, including experience, expertise, clinical studies and RCTs. In many cases, guideline statements reflect a majority (or a consensus) opinion and this may represent the best that the guideline group can come up with at that time. However, an individual clinician (i.e., guideline user) may have more confidence in one expert (i.e., guideline contributor) versus another; such appraisal of colleagues' opinions is common in practice but is almost impossible to convey in conventional guidelines.
- Guidelines are fluid and do change, so you need to keep up on the latest research. The 2012 Surviving Sepsis guidelines, for example, removed a provision about maintaining tight glucose control in the ICU because it was proved to be harmful. Yet, this practice had been advocated in several guidelines.
- Use local knowledge in applying guidelines. In Australia and New Zealand, hospitals did not adopt the Surviving Sepsis guidelines. Outcomes from sepsis seem to be superior in Australia and New Zealand and the approaches recommended in the guidelines did not conform to local practice. Some may argue that even with good baseline outcomes, adopting certain guidelines would provide incremental improvement that although small, may be worthwhile; this assumption is not necessarily true and can be hazardous (e.g. tight glucose control in the ICU).
Are you involved in hospital medicine? Then you should be getting ACP Hospitalist and ACP HospitalistWeekly. Subscribe now.
From the March 25, 2015 edition
- EGDT provided no benefit compared to usual septic shock care
- Using a threshold of 2 SIRS criteria to define severe sepsis may not be adequate, study finds
ACP Career Connection
Looking for a new hospitalist position?
ACP Career Connection can help you find your next job in hospital medicine. Search hospitalist positions nationwide that suit your criteria and preferences. Jobs are posted about two weeks before print publication of Annals of Internal Medicine, ACP Internist, and ACP Hospitalist. Exclusive “Online Direct” opportunities are updated weekly. Check us out online.
ABIM Maintenance of Certification for Hospitalists
Hospital-based internists have the option of maintaining their certification in either Internal Medicine or Internal Medicine with a Focused Practice in Hospital Medicine. Learn more about resources from ACP and the Society for Hospital Medicine to complete both MOC programs.
What will you learn from your Annals Virtual Patient?
Annals Virtual Patients is a unique patient care simulator that mirrors real patient care decisions and consequences. CME Credit and MOC Points are available. Start off with a FREE sample case. Start your journey now.
Internal Medicine Meeting 2015 Live Simulcast!
Unable to attend the meeting this year? On Saturday, May 2, seven sessions will be streamed live from the meeting. Register for the simulcast and earn CME credit after watching each session. Watch it live or download for later viewing.