Is routine measurement of international normalized ratio necessary as part of the investigation of patients with cardiac-type chest pain?

Samuel G.Campbell, Kirk Magee, Ismail Cajee, Simon Field, Michael B.Butler, Christine L.Campbell, Sarah E.Bryson

Charles V.Keating Emergency and Trauma Centre, Queen Elizabeth II HSC, Halifax, Nova Scotia B3H 3A7, Canada

Dear editor,

Chest pain is a frequent complaint of patients presenting to the emergency department (ED), and many of them are referred to the cardiology service for further investigation.At the Charles V.Keating Emergency and Trauma Centre in Halifax, Nova Scotia, Canada,4,800 (6.6%) of the approximately 73,000 patients per year register with a complaint of “chest pain”, and 20% of patients are referred to cardiology.Coagulation studies, specif ically international normalized ratio (INR)frequently part of the “routine” panel of blood tests,are ordered for patients in the ED being investigated or treated for chest pain suspected to be cardiac in nature.Recent calls to examine how much of our practice is likely to benefit patients in any way have led us to question the clinical utility of routine use of these tests.

METHODS

This retrospective ED chart review involved a subanalysis of 1,000 patients referred to cardiology for the investigation of chest pain.The primary study was designed to investigate the use of routine chest X-ray in this population and patients were excluded if they had a requirement for chest X-ray (apart from a history of non-traumatic chest pain), as defined by nine of the ten described by Rothrock et al[1](temperature ≥38 ℃,oxygen saturation ＜90%, respiratory rate ＞24 breaths/minute, hemoptysis, rales, diminished breath sounds, a history of alcohol abuse, tuberculosis, or thromboembolic disease).We included patients with the criterion of “age≥60 years” as age was neither a “disease” nor an abnormal clinical variable, and excluding them would not only lead to the inference that older age is an indicator for“routine” ED screening, but also consider that many of our patients fall into this category and would also decrease the potential for reducing the number of routine ED tests.

For this sub-study investigating the utility of routine INR testing in this group, only patients with INR testing were included.In consideration that it might be reasonable to order INR tests on patients known to be on warfarin,patients that had information in their ED clinical record indicating current warfarin use were further excluded.

Chart audits were conducted according to the recommendations of Gilbert et al[2]with intensive instruction of abstractors into specific and explicit abstraction protocols, defining important variables precisely.Abstractors were trained on a set of “practice”ED medical records before the start of the study, in order to ensure the accuracy of the data gathered and the consistency in which clinical details were recorded.Frequent meetings were held with the abstractors and study coordinators to resolve disputes and review coding rules.Abstractors were blinded as to the hypothesis being tested (i.e., whether the tests were helpful to patient care or not).To reduce the incidence of transcription errors, data were entered directly (concurrent with abstraction) into a computerized data abstraction form developed by the Dalhousie Emergency Database Manager.We have used this process successfully in two previous ED studies.[3,4]

Data abstracted from the ED chart included vital signs, risk factors for ischemic cardiac disease, prior medical history, chest X-ray findings, and clinical f indings.Clinical details and INR results on the database were reviewed independently by four experienced ED faculty (average 25.5-year experience, range 18-37 years).Reviewers were instructed to consider an INR result as “relevant” if they would like to know the particular result when managing a patient not on warfarin therapy with chest pain.

The medical records (beyond the details abstracted from the ED chart) of patients with positive results were examined by one investigator (SGC) to determine if there was a reason for the result that had not been recorded in the ED chart.

RESULTS

Of 1,000 patients identif ied for review in the original study, clinical data were available for 984 (98.4%).A total of 49 patients were excluded because the abstractors had found evidence of current warfarin therapy on the ED chart, leaving 935 patients for review.Of these, INR was measured in 736 (78.7%) patients.Thirteen patients had abnormal INR results (INR ＞1.2).Although no INR f indings were found by any of the reviewers to contribute to the emergency care of these patients, every one of the ten cases (INR ＞1.5) was considered potentially“relevant” by any one of the reviewers; the three cases in which the INR was 1.3 were considered “irrelevant”.

Subsequent depth review of the complete medical records of patients with abnormal results revealed that of the results considered “relevant” in nine cases, the active use of warfarin by the patient had not been recorded in the ED chart, and one patient had a known coagulation anomaly (factor X deficiency), which was not listed as one to be screened for in the abstractors’ protocol.Of the three remaining abnormal results, that were considered“irrelevant” by the reviewers, all three were on other anticoagulants (two dabigatran and one clopidogrel).A flow chart depicting how patients were analyzed is shown in Figure 1.

Figure 1.Flow chart describing how patients were evaluated.

DISCUSSION

In the patient population studied, there were no INR results that led to a change in therapy, and when corrected for incomplete ED recording of current medication and omission of congenital coagulation abnormalities by the abstraction process, no results were unexpected or relevant.Our results were very similar to those of previous investigators.[5,6]

Unnecessary application of medical care is of significant current interest.The Choosing Wisely campaign, released in the USA in 2012[7]and in Canada in 2014,[8]aims to promote conversations between clinicians and patients by helping patients choose care that is supported by evidence, not duplicative of other tests or procedures already received, free from harm, and truly necessary.[9]

The Canadian Association of Emergency Physicians released ten specialty specific recommendations.[10]The basic premise of choosing wisely, however, is far larger than groups of recommendations; it is a call for healthcare providers to consider the risk benefit ratio of each test or procedure that we carry out.Whilst this idea seems intuitive, in many cases interventions have crept into routine use that occur without any thought.In fact in many EDs, panels of biochemical tests are labelled “routine” implying that their use is appropriate for all patients and without which test might be considered somehow deficient.The measuring of indicators of coagulation status is one of such tests.The INR provides a consistent way for reporting what a patient’s prothrombin time (PT) ratio would have been if measured by using the primary World Health Organization International Reference reagent.[11,12]The PT will be prolonged if the concentration of any of the tested factors is 10% or more below normal plasma values.The prolonged PT indicates a def iciency of over 10% below normal plasma values of factors Ⅶ, X, V,prothrombin, or fibrinogen.Causes of the deficiency include vitamin K def iciency, liver failure, disseminated intravascular coagulation (DIC), or the use of vitamin K antagonists, such as warfarin.

Patients for whom it is felt necessary to treat or exclude acute coronary syndrome often end up treatment with anticoagulants (usually fractionated or un-fractionated heparin), and in some cases proceed to cardiac catheterization, which involves invasive vascular manipulation.Presumably these factors have led to the belief that such patients could suffer harm if they had a coagulopathy that their caregivers were unaware of,which could be avoided by routine measurement of their INR.Despite this logic researchers have not been able to support it.[5,6,12-14]In patients with liver disease,there is evidence that an abnormal INR does not expose the patient to the same risk of bleeding as that induced therapeutically.[13]

In our institution, 80% of patients with chest pain are managed by the emergency physician without cardiology consultation.In order to ensure that we were not including chest pain of obviously non-cardiac origin, we limited our study to patients with chest pain subsequently referred to cardiology.It was likely that many patients with chest pain who were not referred to cardiology had workup that included INR testing, making the potential impact of removing INR from “routine practice” far more than we demonstrated.Furthermore, although we chose to focus on this def ined group of patients for whom INR tests were routinely ordered, our findings showed that the “screening” use of these tests were not likely to serve ED patients in other categories who did not have a specif ic emergency indication for their use.We hope that our findings will add to those of others[5,6,12-14]to remind clinicians to pause and consider the potential utility of this test before ref lexively ordering it.

Although guidelines for the ordering of coagulation studies do exist,[12]changing physician practice, is as a rule, challenging.[14]In the specific case of INR testing,restricting the “routine” use of the test, in addition to removing it from recommended panels, might also be facilitated by the fact that the test requires a different(blue top) sodium citrate containing collection tube,which can be stored in a different area from other“routine” tubes, requiring a special trip, and consequently specif ic thought, before the blood is drawn.[12]

Limitations

This study carried the limitations common to all retrospective chart reviews.Although we tried to limit these by a strict chart abstraction process, this involved only a review of the ED chart, resulting in delayed identification of current warfarin use in nine patients where the history was found elsewhere in the patient (non-emergency) record.This finding essentially demonstrated limited history taking and/or documentation in the ED chart, but an unfortunate finding restricted to a very small proportion of cases.A further limitation was that patients were excluded if they had any of nine of the ten Rothrock criteria, which were described to limit unnecessary chest X-ray and not blood tests.While we conceded that this might lead us to miss potential savings by identifying a low utility of INR testing in patients excluded for requiring a chest X-ray,it was unlikely that patients with these criteria would be considered for a “reduced investigation” plan of action,at least at this early stage.

CONCLUSIONS

INR testing is rarely useful in the ED management of patients with cardiac-type chest pain, and its use should be restricted to those in whom coagulation abnormalities might be suspected and for whom knowledge thereof would make a diff erence to patient care.

Funding:None.

Ethical approval:The study was approved by the institutional ethics review board of Nova Scotia Health.

Conflicts of interest:No benefits have been received from a commercial party related directly or indirectly to the study.

Contributors:SGC proposed and wrote the f irst draft.All authors contributed to the design and interpretation of the study an d to further drafts.