Emergency physicians love their autonomy, but adopting the right clinical protocols and “pathways” can result in safer, more efficient care
Dear Director: Our hospital has asked us to work with cardiology and develop a chest pain protocol to decrease some of the low-risk chest pain observation cases. I know there are some EDs that are doing this but this seems like a big change from our daily practice and we have a lot of concerns. What do you think about this?
There are few values more important to the identity of an emergency physician than autonomy. We pride ourselves on the individualized care we provide to each patient based on our bedside clinical judgment. Nothing, we say, should stand between our patients and us. At the same time, we strive to find evidence-based best practices to inform our decisions. Over the past couple of decades, our specialty has produced a number of useful evidence-based pathways to support our care. We all know, and hopefully use, the Ottawa ankle rules, the PECARN head CT rules, and the Canadian or NEXUS c-spine rules. These have allowed us to reduce variation in our care, avoid unnecessary imaging, and provide a defensible basis for our decisions, without confining us to a rigid “cookbook medicine.” These well-known “rules” are a great start, but there are myriad other conditions that would benefit from evidence-based guidelines to optimize our care.
Balanced against evidenced based medicine is our own history and training regarding our approach to chest pain. Probably most of us can remember that one attending from residency who never let anyone be discharged from the ED who spoke the words “chest” and “pain” in close proximity to each other. They’d been burned once and wouldn’t be burned again. If you’ve practiced long enough, you’ve no doubt seen someone present to the ED in cardiac arrest or with a STEMI within months of having a “normal” stress test. These experiences are seared into our thought processes and play a role in our decision-making.
Driving Forces
When implementing a pathway, it’s important to consider the impact and motivation to all the players—the patient, the physician, the ED, and the hospital. There are multiple factors impacting the development and implementation of more clinical pathways. The Ottawa ankle rule is great, but if a patient comes to the ED and really wants an x-ray, how many physicians want to spend the extra time with the patient explaining why they don’t need an x-ray, and perhaps risk a lower patient sat survey score? While there are some patients who will happily leave the ED without the x-ray, many would rather wait the hour “to be 100% sure.” There is low risk and cost to the patient by staying, not a big impact to the ED, and it makes the physician’s job easier to just click on the x-ray. Compare that to using the PECARN head CT rule where we save the child radiation exposure, appeal to the parents about avoiding an increased risk of cancer, save the parents from a higher cost test, and likely save the ED significant time. While we haven’t studied it, we suspect there is greater buy-in from the emergency medicine community about implementing PECARN than the Ottawa ankle rules. It’s about aligning the motivations and risk of the provider and patient.
Chest pain is, as we all know, one of the most common chief complaints in the ED, accounting for 1 in every 18 patients we see. It’s a high-risk condition, and accounts for more malpractice suits, and higher average payouts, then any other complaint. And it’s not an easy one for us to manage alone. We’re often caught between the conflicting interests of our own concern for defensible practice, reluctant hospitalists, and a lack of helpful guidelines on which to base our practice.
Now throw on the financial and time pressures of hospitals that may want to reduce observation cases because of low reimbursement, cardiologists who would prefer to stress these patients as outpatients, and the potential impact to ED flow. For most of us without adjacent observation units or who don’t practice in Chest Pain ER environments, the patients who go upstairs for an obs/admission typically stay longer in the ED, thus creating back ups, and fill inpatient beds, which could exacerbate an ED boarding situation.
A recent study in Health Affairs found a greater than 6-fold difference in the chest pain risk-adjusted admission rate between EDs in the 25th and 75th percentiles. That’s nearly twice as much variation as we have in soft tissue infections or asthma, four times as much variation as for pneumonia or CHF, and six times the variation for renal failure or sepsis. Nationally, we’re spending $10 billion annually on these rule-out admissions, but the same study also showed that patients admitted for chest pain have the lowest inpatient mortality rate of any common admission diagnosis, by a wide margin. Patients admitted for non-septic UTIs are dying 15 times more often; septic patients 294 times more often.
In our own ED group, we found a huge amount of variation in our practice patterns even within the same ED. In an online survey of our docs we posed a detailed hypothetical case scenario of a low-risk chest pain patient with prolonged pain and asked our docs how they would usually approach this patient in their daily practice. Amazingly, we were split almost evenly between discharging him (with outpatient follow-up), bringing him into the hospital, or keeping him in the ED for a second troponin. So, exact same patient, same ED, and what determines his care is which one of us happens to pick up his chart. This doesn’t mean that we’re bad doctors, and there isn’t only one “right answer,” but this level of variation is not sustainable, and we need to find ways to help ourselves do better.
One Solution
Just like each patient encounter is in many ways unique, each hospital has unique features. Whether it’s a chest pain unit, bedside makers, an ED-controlled Obs unit, or none of the above, there are existing protocols that will serve as a starting point for your development of a pathway to safely and efficiently manage your low risk chest pain patients.
Posed with similar questions by our hospital and cardiologists, we formed a committee of our ED docs to identify conditions where there’s good evidence to guide our clinical care, and then created clinical pathways to support our decision making processes and provide more consistent, higher value care to our patients. We use the term “pathways” intentionally: these aren’t hard and fast “rules” that have no exceptions. We still need to use our individual clinical judgment, but we can at least all be using the same playbook as our default course.
In the past, trying to create a rational approach to chest pain has been hampered by a lack of clear guidelines and evidence, but this is rapidly changing. The 2014 AHA/ACC NSTEMI guidelines are still far from perfect, but are an improvement over the prior revisions. There have been dozens of new studies, in the ED and cardiology literature, assessing improved approaches to low-risk chest pain patients in the ED. Perhaps most importantly, many EDs around the country, including a number of leading academic centers, have been implementing chest pain pathways, helping to define the standard of care.
I’ve had cardiologists push TIMI Scoring on my EDs multiple times over the past 15 years. Although risk stratification is a key element of any chest pain pathway, the TIMI score includes unintuitive elements (having taken aspirin in the past week hardly seems like a logical risk factor), is difficult to remember, only captures a small percentage of patients into the lowest risk group, and its negative predictive value is less than ideal. As an alternative, the HEART score is another validated risk stratification aid that has recently been gaining popularity. In contrast to TIMI Scores, the HEART score is easy to use and remember (it’s a mnemonic device), logically incorporates our clinical judgment, captures a sizeable proportion of patients into the low-risk category, and has an excellent negative predictive value. There’s a growing body of literature to support outpatient management of patients with a HEART score of 0-3, and this is rapidly becoming the standard of care at more and more hospitals across the country. If a patient is low risk by HEART, has a non-ischemic EKG and negative troponin six or more hours after onset of pain, or a negative three-hour delta for more recent onset, she is safe to complete her workup as an outpatient, and avoid a wasteful hospital stay.
The catch is ensuring that urgent outpatient cardiology follow-up is a reliable option. This can be a challenge, especially for uninsured patients, but as healthcare payment options are evolving, incentives are aligning to push everybody to make this work. Guaranteeing an urgent cardiology appointment for a low-risk patient helps the flow of the ED, saves the patient an unnecessary night in the hospital, avoids a money-losing observation-status admission for the hospital, and brings outpatient business to the cardiologists. For these reasons, as the incentives change for hospitals, some are starting to subsidize these outpatient visits. After all, an outpatient appointment with a cardiologist costs far less than a night in the hospital if the hospital isn’t getting any incremental revenue, such as for uninsured patients, or in capitated payment systems. Clearly, implementing a chest pain pathway requires cooperation with the cardiologists, hospitalists, and administration; so expect to spend a few months consensus building before you can roll out a pathway.
Conclusion
I had a trauma attending as a med student who said he loved trauma because taking care of patients was like following a recipe. At the time, occurring right after a medicine rotation, I thought it should involve more thought and individual care. Now, I’ve become a huge fan of pathways. It allows me to practice evidenced based medicine, document appropriately, protect me from malpractice, and provide more efficient care to the patients. Regarding chest pain, most patients want to go home and it’s usually me, the physician, who’s reluctant to discharge them.
There are no “off the shelf” solutions for tackling chest pain that are ready to roll out in your ED tomorrow. Improving chest pain practice will require work to create a pathway that works in a particular hospital and to set up a reliable mechanism for outpatient follow-up. Most importantly is to get buy-in from your fellow EM docs, as well as from the cardiologists, hospitalists, and administrators. Don’t forget to talk to your PMDs who will do some of the follow up when you’re consensus building. If done right, we can provide a more consistent, safer, more defensible, and more efficient care to our patients.
1 Comment
I’m getting rather tired of “chest pain” being discussed as if our only job is to rule out MI. I’ve seen too many cases of an MI being ruled out, only to have the patient die of a missed PE or thoracic aortic dissection (TAD). Considering it’s rarity, TAD is a more common cause of malpractice lawsuits than missing an MI. I suspect that the low PERCENTAGE of chest pain patients who die in hospital includes a high NUMBER of patients who die not of an MI but of another cause of chest pain that was missed because of our hyper-focus on MI.