Diagnostic Performance of Emergency Physician Gestalt for Predicting Acute Appendicitis in Patients Age 5 to 20 Years

Predictive values of indirect ultrasound signs for low risk of acute appendicitis in paediatric patients without visualisation of the appendix on ultrasound

BACKGROUND AND OBJECTIVES

The ability to rule appendicitis in or out using ultrasound is limited by studies where the appendix is not visualised. We determined whether the absence of indirect ultrasound signs can rule out appendicitis in children undergoing a radiology-performed ultrasound in which the appendix is not visualised METHODS: This was a single-centre retrospective observational study of patients aged 3-13 with a clinical suspicion of acute appendicitis evaluated in a Paediatric Emergency Department in Spain from 1 January 2013 to 31 December 2019. For those patients who had formal ultrasound, direct and indirect findings of ultrasound were abstracted from the ultrasound report. The surgical pathology report was established as the gold standard in patients who underwent an appendectomy. In those who did not, appendicitis was considered not to be present if there was no evidence in their charts that they had undergone an appendectomy or conservative therapy for appendicitis during the episode. The main outcome variable was the diagnosis of acute appendicitis. For patients undergoing ultrasound, the independent association of each indirect ultrasound sign with the diagnosis of appendicitis in patients without a visualised appendix was analysed using logistic regression.

RESULTS

We included 1756 encounters from 1609 different episodes. Median age at the first visit of each episode was 10.1 years (IQR, 7.7-11.9) and 921 (57.2%) patients were men. There were 730 (41.6%) encounters with an Alvarado score ≤3, 695 (39.6%) with a score 4-6 and 331 (18.9%) with a score ≥7. Appendicitis was diagnosed in 293 (17.8%) episodes. Ultrasonography was performed in 1115 (61.6%) encounters, with a visualised appendix in 592 (53.1%).The ultrasound findings independently associated with appendicitis in patients without a visualised appendix were the presence of free intra-abdominal fluid in a small quantity (OR:5.0 (95% CI 1.7 to 14.6)) or in an abundant quantity (OR:30.9 (95% CI 3.8 to 252.7)) and inflammation of the peri-appendiceal fat (OR:7.2 (95% CI 1.4 to 38.0)). The absence of free fluid and inflammation of the peri-appendiceal fat ruled out acute appendicitis in patients with an Alvarado score <7 with a sensitivity of 84.6% (95% CI 57.8 to 95.7) and a negative predictive value of 99.4% (95% CI 97.8 to 99.8).

CONCLUSIONS

Patients with an Alvarado score <7 and without a visualised appendix on ultrasound but who lack free fluid and inflammation of the peri-appendiceal fat are at very low risk of acute appendicitis.

Diagnostic accuracy of doctors at the emergency department and radiologists in differentiating between complicated and uncomplicated acute appendicitis

PURPOSE

To determine the accuracy of final judgements of doctors at the emergency department (ED) and radiologists to differentiate between complicated and uncomplicated acute appendicitis, because these have different treatment options.

METHODS

This prospective, multicenter study included adult patients with imaging-confirmed acute appendicitis, operated with intention to appendectomy. Both doctors at ED and radiologists assessed appendicitis severity as a final judgement of "uncomplicated" or "complicated" appendicitis. Doctors at ED integrated clinical, laboratory, and imaging findings. Radiologists relied solely on imaging findings. Outcomes were accuracy of these judgements for diagnosis of complicated appendicitis compared to the reference standard by an adjudication committee.

RESULTS

After imaging, 1070 patients with confirmed acute appendicitis were included. Doctors at ED accurately labelled 656 of 701 (93.6%) patients with true uncomplicated appendicitis as uncomplicated, and 163 of 369 (44.2%) patients with true complicated appendicitis were labelled as complicated. Sensitivity, specificity, and positive and negative predictive values (PPV and NPV) for complicated appendicitis were 44.2%, 93.6%, and 78.4% and 76.1%, respectively. Comparable accuracy was found for the radiologist's assessment in 941 patients, with true positive rates of 92.2% (581 of 630 patients) for uncomplicated appendicitis and 46.6% (145 of 311 patients) for complicated appendicitis.

CONCLUSION

More than half of all patients with true complicated appendicitis is incorrectly classified as uncomplicated appendicitis according to the judgements of doctors at ED, integrating clinical, laboratory, and imaging results, and of radiologists assessing diagnostic imaging. These judgements are thereby not sufficiently reliable in ruling out complicated appendicitis.

Outcomes of patients enrolled in a prospective and randomized trial on basis of gestalt assessment or ABC score

BACKGROUND

The Pragmatic Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial rapidly enrolled patients based on an Assessment of Blood Consumption (ABC) ≥ 2 score, or physician gestalt (PG) when ABC score was <2. The objective of this study was to describe what patients were enrolled by the two methods and whether patient outcomes differed based on these enrollments. We hypothesized that there would be no differences in outcomes based on whether patients were enrolled via ABC score or PG.

METHODS

Patients were enrolled with an ABC ≥ 2 or by PG when ABC was <2 by the attending trauma surgeon. We compared 1-hour, 3-hour, 6-hour, 12-hour, 18-hour, and 24-hour mortality, 30-day mortality, time to hemostasis, emergent surgical or interventional radiology procedure and the proportion of patients who required either >10 units of blood in 24 hours or >3 units in 1 hour.

RESULTS

Of 680 patients, 438 (64%) were enrolled on the basis of an ABC score ≥2 and 242 (36% by PG when the ABC score was <2). Patients enrolled by PG were older (median, 44; interquartile range [IQR], 28-59; p < 0.001), more likely to be White (70.3% vs. 60.3%, p = 0.014), and more likely to have been injured by blunt mechanisms (77.3% vs. 37.2%, p < 0.001). They were also less hypotensive and less tachycardic than patients enrolled by ABC score (both p < 0.001). The groups had similar Injury Severity Scores in the ABC ≥ 2 and PG groups (26 and 27, respectively) and were equally represented (49.1% and 50.8%, respectively) in the 1:1:1 treatment arm. There were no significant differences between the ABC score and PG groups for mortality at any point. Time to hemostasis (108 for patients enrolled on basis of Gestalt, vs. 100 minutes for patients enrolled on basis of ABC score), and the proportion of patients requiring a massive transfusion (>10 units/24 hours) (44.2% vs. 47.3%), or meeting the critical administration threshold (>3 unit/1 hour) (84.7% vs. 89.5%) were similar ( p = 0.071).

CONCLUSION

Early identification of trauma patients likely to require a massive transfusion is important for clinical care, resource use, and selection of patients for clinical trials. Patients enrolled in the PROPPR trial based on PG when the ABC score was <2 represented 36% of the patients and had identical outcomes to those enrolled on the basis of an ABC score of ≥2.

LEVEL OF EVIDENCE

Prognostic and Epidemiological; Level III.

Gestalt clinical severity score (GCSS) as a predictor of patient severity of illness or injury

OBJECTIVE

To determine if clinical judgement is accurate to predict the severity of injury or illness, and can be used at patient arrival when other formal scoring systems are not yet available.

DESIGN

A multicenter pilot study using a prospective observational convenience sample of patients arriving by EMS to the emergency department (ED) or Trauma Center.

SETTING

Two urban, Level 1 trauma centers at academic tertiary care hospitals.

PATIENTS

Medical and trauma patients age 18 and older transported by EMS (N = 216). Exclusion criteria (prior to arrival): intubation, assisted ventilation (BVM or NPPV), CPR in progress, prisoners, or previously present motor or speech deficits.

MEASUREMENTS

Completion of a novel 15-point scale of Verbal, Motor, and Facial Expression within 1-2 min of arrival by a clinician outside of the treatment team. Primary endpoint was the immediate disposition from the ED or Trauma Center: Home, Brief Observation (<24 h), Admission to Floor, ICU (OR and IR as surrogates since these patients ultimately go to the ICU), or Morgue.

RESULTS

Univariate analysis revealed a strong, positive monotonic correlation between GCSS and disposition (Rho = 0.693, p < .0001). Multivariable logistic regression revealed the "best" model included GCSS and age (group 18-44 years old versus all the other age groups) (p < .0001). There was a 156% increase in the odds of being discharged home (versus being admitted) for a one-unit increase in GCSS (OR = 2.56, 95% CI 1.94, 3.37).

CONCLUSIONS

Physicians can make accurate predictions of severity of injury and illness using a gestalt method and the scoring system we have developed as patient disposition correlates well with GCSS score. GCSS is most accurate with the 18-44 age group.

Physician gestalt for emergency department triage: A prospective videotaped study

OBJECTIVE

Appropriate triage in patients presenting to the emergency department (ED) is often challenging. Little is known about the role of physician gestalt in ED triage. We aimed to compare the accuracy of emergency physician gestalt against the currently used computerized triage process.

METHODS

We conducted a prospective observational study in the ED at an academic medical center. Adult patients aged ≥20 years were included and underwent a standard triage protocol. The patients underwent system-based triage using the computerized software the Taiwan Triage and Acuity Scale. The entire triage process was recorded, and triage data were collected. Five physician raters provided triage levels (physician-based) according to their perceived urgency after reviewing videos. The primary outcome was hospital admission. The secondary outcomes were ED length of stay (EDLOS) and charges.

RESULTS

In total, 656 patients were recruited (mean age 52 years, 50% male). The median system-based triage level was 3. By contrast, the median physician-based triage level was 4. The physician raters tended to provide lower triage levels than the system, with an average difference of 1. There was modest concordance between the two triage methods (correlation coefficient 0.30), with a weighted kappa coefficient of 0.18. The area under the receiver operating curve for the system- and physician-based triage in predicting hospital admission were similar (0.635 vs. 0.631, p = 0.896). Attending physicians appeared to have better performance than residents in predicting admission. The variation explained (R² ) in EDLOS and charges were similar between the two triage methods (R²  = 3% for EDLOS, 7%-9% for charges).

CONCLUSIONS

Emergency physician gestalt for triage showed similar performance to a computerized system; however, physicians redistributed patients to lower triage levels. Physician gestalt has advantages for identifying low-risk patients. This approach may avoid undue time pressure for health care providers and promote rapid discharge.

Factors affecting utilization of CT scan following ultrasound evaluation of suspected appendicitis

CONTEXT

When evaluating a pediatric patient in the emergency department for suspected appendicitis, a provider is often faced with the dilemma of deciding if a computed tomography (CT) scan is warranted when-as is most often the case-ultrasound results do not yield a definitive diagnosis. The potential risks of radiation must be weighed against numerous aspects of a patient's background, physical exam, and already-obtained workup.

OBJECTIVES

This study aims to aid in future decision making of providers in their evaluation of patients with suspected appendicitis, to help facilitate a more comprehensive answer to the "next-steps" in the question of equivocal ultrasound, and to create a pathway utilizing lab results, physical exam findings, and pertinent positives and negatives in patient history to facilitate a more objective decision-making process for ordering a CT scan.

METHODS

A retrospective chart review was performed for patients who were evaluated for possible appendicitis at our institution between October 2016 and October 2017 and whose ultrasound results were equivocal. We determined what factors led to increased frequency of obtaining CT scans.

RESULTS

Statistical analysis showed that several factors were independently associated with the increased likelihood of having a CT scan performed. When considered in combination with the other parameters, it was found that older children, males, report of right lower quadrant (RLQ) pain, and presence of RLQ tenderness on physical exam were all associated with a significantly higher likelihood of having a CT scan performed.

CONCLUSIONS

When combined with present algorithms and protocols already in use, this information can assist the provider in making prudent decisions for their patients with the potential for reduced provider bias.

Decision Making: Healthy Heuristics and Betraying Biases

Critical care settings are unpredictable, dynamic environments where clinicians face high decision density in suboptimal conditions (stress, time constraints, competing priorities). Experts have described two systems of human decision making: one fast and intuitive; the other slow and methodical. Heuristics, or mental shortcuts, a key feature of intuitive reasoning, are often accurate, applied instinctively, and essential for efficient diagnostic decision making. Heuristics are also prone to failures, or cognitive biases, which can lead to diagnostic errors. A variety of strategies have been proposed to mitigate biases; however, current understanding of such interventions to optimize diagnostic safety is still incomplete.

Prediction of patient disposition: comparison of computer and human approaches and a proposed synthesis

OBJECTIVE

To compare the accuracy of computer versus physician predictions of hospitalization and to explore the potential synergies of hybrid physician-computer models.

MATERIALS AND METHODS

A single-center prospective observational study in a tertiary pediatric hospital in Boston, Massachusetts, United States. Nine emergency department (ED) attending physicians participated in the study. Physicians predicted the likelihood of admission for patients in the ED whose hospitalization disposition had not yet been decided. In parallel, a random-forest computer model was developed to predict hospitalizations from the ED, based on data available within the first hour of the ED encounter. The model was tested on the same cohort of patients evaluated by the participating physicians.

RESULTS

198 pediatric patients were considered for inclusion. Six patients were excluded due to incomplete or erroneous physician forms. Of the 192 included patients, 54 (28%) were admitted and 138 (72%) were discharged. The positive predictive value for the prediction of admission was 66% for the clinicians, 73% for the computer model, and 86% for a hybrid model combining the two. To predict admission, physicians relied more heavily on the clinical appearance of the patient, while the computer model relied more heavily on technical data-driven features, such as the rate of prior admissions or distance traveled to hospital.

DISCUSSION

Computer-generated predictions of patient disposition were more accurate than clinician-generated predictions. A hybrid prediction model improved accuracy over both individual predictions, highlighting the complementary and synergistic effects of both approaches.

CONCLUSION

The integration of computer and clinician predictions can yield improved predictive performance.

Accuracy of Emergency Department Clinical Findings for Diagnosis of Coronavirus Disease 2019

Study objective

We seek to describe the medical history and clinical findings of patients attending the emergency department (ED) with suspected coronavirus disease 2019 (COVID-19) and estimate the diagnostic accuracy of patients’ characteristics for predicting COVID-19.

Methods

We prospectively enrolled all patients tested for severe acute respiratory syndrome coronavirus 2 by reverse-transcriptase polymerase chain reaction in our ED from March 9, 2020, to April 4, 2020. We abstracted medical history, physical examination findings, and the clinical probability of COVID-19 (low, moderate, and high) rated by emergency physicians, depending on their clinical judgment. We assessed diagnostic accuracy of these characteristics for COVID-19 by calculating positive and negative likelihood ratios.

Results

We included 391 patients, of whom 225 had positive test results for severe acute respiratory syndrome coronavirus 2. Reverse-transcriptase polymerase chain reaction result was more likely to be negative when the emergency physician thought that clinical probability was low, and more likely to be positive when he or she thought that it was high. Patient-reported anosmia and the presence of bilateral B lines on lung ultrasonography had the highest positive likelihood ratio for the diagnosis of COVID-19 (7.58, 95% confidence interval [CI] 2.36 to 24.36; and 7.09, 95% CI 2.77 to 18.12, respectively). The absence of a high clinical probability determined by the emergency physician and the absence of bilateral B lines on lung ultrasonography had the lowest negative likelihood ratio for the diagnosis of COVID-19 (0.33, 95% CI 0.25 to 0.43; and 0.26, 95% CI 0.15 to 0.45, respectively).

Conclusion

Anosmia, emergency physician estimate of high clinical probability, and bilateral B lines on lung ultrasonography increased the likelihood of identifying COVID-19 in patients presenting to the ED.