Month: December 2019

The Clinical Impact of SAH Decision Rules

The Ottawa Subarachnoid Hemorrhage Rule has been around a long time now, dating back to 2013. The “six hour CT” rule has been around even longer, dating back even to 2011. They’ve become entwined in at least the discussion around the evaluation of SAH, if not clinical practice.

… but are they actually useful?

This is the “before and after” study from Perry and Stiell (not to be confused with Penn and Teller), in which the practice of Canadian physicians was examined around the time these rules were under development and in publication. These authors gathered data on patients presenting with atraumatic headache spanning the time period between 2011 and 2016, looking at resource utilization and missed SAH before and after adoption of both the Ottawa SAH Rule and the 6-hour CT Rule. Specifically, practicing clinicians were instructed not to use decision rules for the basis of patient care until June 2013, at which point clinicians were actively encouraged to do so.

The basic findings:

  • The Ottawa SAH Rule doesn’t change much.
  • The 6-hour CT Rule probably reduces downstream lumbar puncture/CTA.

Again, with concern for generalizability, a full 5.1% of their qualifying atraumatic headaches were diagnosed with SAH across the study period. The rate of investigation of these patients remained high, about 88%, regardless of study period – and regardless whether the Ottawa Rule criteria were met. However, for patients presenting within 6 hours of headache onset, the rate of subsequent LP dropped from 31.3% to 15.1%. The Ottawa SAH Rule showed its expected specificity of about 12%, and, therefore, was 100% sensitive. The 6-hour CT Rule “missed” 5 of 111 patients, however, for various reasons – one radiology misread, a false-positive owing to profound anemia, a non-aneurysmal SAH from dural vein fistula, and two cases of false-positive LPs meeting their study criteria for false-negative CT.

A long story made short, 1) keep using the 6-hour CT rule with the caveat of known potential confounders to visible blood (anemia); 2) the Ottawa Rule is only clinical useful as a one-way decision instrument owing to its poor positive likelihood ratio.

“Prospective Implementation of the Ottawa Subarachnoid Hemorrhage Rule and 6-Hour Computed Tomography Rule”

https://www.ncbi.nlm.nih.gov/pubmed/31805846

Sepsis Alerts Save Lives!

Not doctors, of course – the alerts.

This is one of those “we had to do it, so we studied it” sorts of evaluations because, as most of us have experienced, the decision to implement the sepsis alerts is not always driven by pent-up clinician demand.

The authors describe this as sort of “natural experiment”, where a phased or stepped roll-out allows for some presumption of control for unmeasured cultural and process confounders limiting pre-/post- studies. In this case, the decision was made to implement the “St John Sepsis Algorithm” developed by Cerner. This algorithm is composed of two alerts – one somewhat SIRS- or inflammation-based for “suspicion of sepsis”, and one with organ dysfunction for “suspicion of severe sepsis”. The “phased” part of the roll-out involved turning on the alerts first in the acute inpatient wards, then the Emergency Department, and then the specialty wards. Prior to being activated, however, the alert algorithm ran “silently” to create the comparison group of those for whom an alert would have been triggered.

The short summary:

  • In their inpatient wards, mortality among patients meeting alert criteria decreased from 6.4% to 5.1%.
  • In their Emergency Department, admitted patients meeting alert criteria were less likely to have a ≥7 day inpatient length-of-stay.
  • In their Emergency Departments, antibiotic administration of patients meeting alert criteria within 1 hour of the alert firing increased from 36.9% to 44.7%.

There are major problems here, of course, both intrinsic to their study design and otherwise. While it is a “multisite” study, there are only two hospitals involved. The “phased” implementation not the typical different-hospitals-at-different-times, but within each hospital. They report inpatient mortality changes without actually reporting any changes in clinician behavior between the pre- and post- phases, i.e., what did clinicians actually do in response to the alerts? Then, they look at timely antibiotic administration, but they do not look at general antibiotic volume or the various unintended consequences potentially associated with this alert. Did admission rates increase? Did percentages of discharged patients receiving intravenous antibiotics increase? Did clostridium difficle infection rates increase?

Absent the funding and infrastructure to better prospectively study these sorts of interventions, these “natural experiments” can be useful evidence. However, these authors do not seem to have taken an expansive enough view of their data with which to fully support an unquestioned conclusion of benefit to the alert intervention.

“Evaluating a digital sepsis alert in a London multisite hospital network: a natural experiment using electronic health record data”

https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocz186/5607431

YEARS, But Wells

It’s the Monday after Thanksgiving, so it’s time to turn the brains back on – and notice an oddly robust cultivation of articles worthy of comment dropped just before the holiday. This is the first, the “Pulmonary Embolism Graduated D-Dimer” (PEGeD) study, a rather obtuse name for what is effectively a pretest likelihood-adjusted implementation of D-dimer. This is also, effectively, what was done with the YEARS protocol – so, what’s new?

In this iteration of the concept, the authors use the Wells score, stratifying patients to either low, moderate, or high probability. A few published work-up algorithms describe pathways of care in which low probability leads to PERC, while the higher-risk cohort undergoes D-dimer testing or directly to CTPA to rule out PE. In this algorithm, those with “low” probability still undergo D-dimer testing – but with a cut-off threshold of 1000 ng/mL warranting advanced imaging. The primary outcome was symptomatic, objectively-verified venous thromboembolism, including PE and deep venous thrombosis, at 90 days.

These authors enrolled a cohort of 2,017 patients between 2015 and 2018, with 1,752 in the “low” probability cohort, 218 “moderate”, and 47 “high”. Overall prevalence of PE on initial testing was 7% and advanced imaging was performed on 34%, for an imaging yield of 24%. The general finding of most importance to the practicing clinician is their observation that 1,285 low-risk patients had D-dimer <1000 ng/mL and 40 moderate-risk patients had D-dimer <500 ng/mL and none had VTE detected at 90 day follow-up. Helpfully, these authors even compare their yield directly to the YEARS protocol – and find about 40 fewer patients would have undergone imaging with PEGeD than YEARS, which makes it basically a wash. They also compare their strategy to an age-adjusted D-dimer, which is a bit odd, considering they are not competing strategies, but synergistic.

The idea of pretest-adjusted D-dimer has been around a very long time, dating back to at least 2012. There’s nothing magical about a cut-off of 1000 ng/mL other than Round Numbers, but it is a serendipitously reasonable starting point for this approach. The real elephant in the room, however, is there were only 87 PEs in their low-risk cohort, for a prevalence of 4.9%. This may yet even over-represent the prevalence of PE in community practice in certain settings (read: the United States). Considering the accepted miss rate for PE is considered to be at least 1%, owing to the likelihood of false-positives and harms from anticoagulation, it is likely an even more aggressive cut-off or imaging-elimination strategy should be pursued.

However, I certainly do not want to minimize this work – adding good, prospective data pushing imaging stewardship is of great importance, whatever minor shortcomings might be observed. At the very least, please considering using PEGeD or YEARS as the basis for your imaging algorithm – and add age-adjusted D-dimer on top for even better reductions in unnecessary imaging.

“Diagnosis of Pulmonary Embolism with d-Dimer Adjusted to Clinical Probability”
https://www.nejm.org/doi/full/10.1056/NEJMoa1909159