Colin O’Rourke

The Wisdom of Crowds of Doctors: Their Average Predictions Outperform Their Individual Ones

Background. Evidence suggests that the average prediction across groups is more accurate than for individuals. Our goals were therefore to investigate accuracy of the average predictions for groups of clinicians and to compare this accuracy with a published statistical prediction model. Methods. Twenty-four expert clinicians attending an advisory board meeting were asked to make predictions for 25 patients from a research registry regarding the probability of having a positive bone scan 1 year from today if left untreated. Comparisons were made between the accuracy of average responses and that of an appropriate previously published statistical prediction model. Results. This study suggests that the mean of the clinicians’ predictions can quickly approach the accuracy of the best clinician using as few as 5 clinicians. When all 24 clinicians’ predictions were averaged, the concordance index reached 0.750, still far below that of the published statistical model with 0.812. Conclusions. Averaging clinician predictions may have merit over individual clinician predictions but still not reasonably replace a carefully built statistical model. However, averaging clinician predictions could prove helpful in situations where statistical models do not yet exist or where existing models are inadequate.

Statistical Comparison of the Pediatric Versus Adult IKDC Subjective Knee Evaluation Form in Adolescents

Background: The International Knee Documentation Committee (IKDC) Subjective Knee Evaluation Form is a patient-reported outcome with adult (1998) and pediatric (2011) versions validated at different ages. Prior longitudinal studies of patients aged 13 to 17 years who tore their anterior cruciate ligament (ACL) have used the only available adult IKDC, whereas currently the pediatric IKDC is the accepted form of choice. Purpose/Hypothesis: This study compared the adult and pediatric IKDC forms and tested whether the differences were clinically significant. The hypothesis was that the pediatric and adult IKDC questionnaires would show no clinically significant differences in score when completed by patients aged 13 to 17 years. Study Design: Cohort study (diagnosis); Level of evidence, 2. Methods: A total of 100 participants aged 13 to 17 years with knee injuries were split into 2 groups by use of simple randomization. One group answered the adult IKDC form first and then the pediatric form. The second group answered the pediatric IKDC form first and then the adult form. A 10-minute break was given between form administrations to prevent rote repetition of answers. Study design was based on established methods to compare 2 forms of patient-reported outcomes. A 5-point threshold for clinical significance was set below previously published minimum clinically important differences for the adult IKDC. Paired t tests were used to test both differences and equivalence between scores. By ordinary least-squares models, scores were modeled to predict adult scores given certain pediatric scores and vice versa. Results: Comparison between adult and pediatric IKDC scores showed a statistically significant difference of 1.5 points; however, the 95% CI (0.3-2.6) fell below the threshold of 5 points set for clinical significance. Further equivalence testing showed the 95% CI (0.5-2.4) between adult and pediatric scores being within the defined 5-point equivalence region. The scores were highly correlated, with a linear relationship (R2 = 92%). Conclusion: There was no clinically significant difference between the pediatric and adult IKDC form scores in adolescents aged 13 to 17 years. This result allows use of whichever form is most practical for long-term tracking of patients. A simple linear equation can convert one form into the other. If the adult questionnaire is used at this age, it can be consistently used during follow-up.

Safety of Large-volume, Same-day Oral Bowel Preparations During Deep Sedation: A Prospective Observational Study

BACKGROUND: Colonoscopy quality is directly related to the bowel preparation. It is well established that bowel preparations are improved when at least part of the laxative is ingested on the day of the procedure. However, there is concern that this can result in higher gastric residual volumes (GRV) and increase the risk of pulmonary aspiration. The aim of this study is to evaluate GRV and gastric pH in patients who received day-before bowel preparation versus those ingesting their laxative on the day of colonoscopy under anesthesiologist-directed propofol deep sedation. METHODS: This is a prospective observational study for patients undergoing same-day upper endoscopy and colonoscopy. All included patients had large-volume polyethylene glycol lavage preparation and received propofol sedation. Gastric fluid was collected during the upper endoscopy for volume and pH measurement. RESULTS: The study included 428 patients with 56% receiving same-day laxative preparation and the remainder evening-before preparation. Mean ± SD GRV was 18.1 ± 10.2 mL, 16.3 ± 16.5 mL in each of these preparation groups, respectively (P = .69). GRV ≥ 25 mL or higher than expected GRV adjusted by weight (0.4 mL/kg) were also not different among the study groups (P = .90 and P = .87, respectively). Evaluating GRV based on time since last ingestion of preparation (3–5, 5–7, >7 hours) did not result in any differences (P = .56). Gastric pH was also similar between the bowel preparation groups (P = .23), with mean ± SD of 2.5 ± 1.4 for evening-before and 2.5 ± 1.3 for the same-day preparation. There were more inadequate bowel preparations in day before bowel preparations (P = .001). CONCLUSIONS: A large-volume bowel preparation regimen finished on the day of colonoscopy as close as 3 hours before the procedure results in no increase in GRV or decrease in gastric pH.

Erratum to: Fertility in Women of Reproductive Age After Breast Cancer Treatment: Practice Patterns and Outcomes

Background Breast cancer is the most frequently occurring cancer in women of reproductive age, and systemic treatments may adversely affect childbearing plans. Use of assisted reproductive technologies and therapies for ovarian protection improve fertility prospects. We evaluated whether patients had a documented fertility discussion (FD) with their oncology physician prior to therapy, what options were chosen, and if pregnancy was achieved.

No Clinically Significant Difference Between Adult and Pediatric IKDC Subjective Knee Evaluation Scores in Adults

Background: Two versions of the International Knee Documentation Committee (IKDC) Subjective Knee Evaluation form currently exist: the original version (1999) and a recently modified pediatric-specific version (2011). Comparison of the pediatric IKDC with the adult version in the adult population may reveal that either version could be used longitudinally. Hypothesis: We hypothesize that the scores for the adult IKDC and pediatric IKDC will not be clinically different among adult patients aged 18 to 50 years. Study Design: Randomized crossover study design. Level of Evidence: Level 2. Methods: The study consisted of 100 participants, aged 18 to 50 years, who presented to orthopaedic outpatient clinics with knee problems. All participants completed both adult and pediatric versions of the IKDC in random order with a 10-minute break in between. We used a paired t test to test for a difference between the scores and a Welch’s 2-sample t test to test for equivalence. A least-squares regression model was used to model adult scores as a function of pediatric scores, and vice versa. Results: A paired t test revealed a statistically significant 1.6-point difference between the mean adult and pediatric scores. However, the 95% confidence interval (0.54-2.66) for this difference did not exceed our a priori threshold of 5 points, indicating that this difference was not clinically important. Equivalence testing with an equivalence region of 5 points further supported this finding. The adult and pediatric scores had a linear relationship and were highly correlated with an R2 of 92.6%. Conclusion: There is no clinically relevant difference between the scores of the adult and pediatric IKDC forms in adults, aged 18 to 50 years, with knee conditions. Clinical Relevance: Either form, adult or pediatric, of the IKDC can be used in this population for longitudinal studies. If the pediatric version is administered in adolescence, it can be used for follow-up into adulthood.

A Current Assessment of Diversity Characteristics and Perceptions of Their Importance in the Surgical Workforce

IntroductionDiversity in the workforce is vital to successful businesses. Healthcare in general has suffered from a lack of cultural competence, which is the ability to successfully interact with individuals from diverse backgrounds. In order to eliminate discrimination and build a diverse workforce, physicians’ perceptions and importance of diversity need to be measured.MethodsA 25-item, anonymous, online questionnaire was created, and a cross-sectional survey was performed. The instrument consisted of demographic and Likert-style questions which attempted to determine the participants’ perceptions of the current level of diversity in their specialty and their perceived importance of particular diversity categories.ResultsOver 1,000 responses were received from US-based physicians across all specialties and levels of training. Statistically significant differences existed between surgical and nonsurgical specialties with regard to gender, prior work experience, and political identity. In the surgical workforce, there is significant perceived homogeneity regarding gender/sexual identity. Surgical respondents also deemed gender/sexual identity diversity to be less important than respondents from medical specialties.ConclusionSurgeons and surgical trainees are less diverse than their medical colleagues, both by demographics and self-acknowledgement. The long-term impact and potential barriers to resolve these differences in diversity require further investigation.

Can polyp detection rate be used prospectively as a marker of adenoma detection rate

BackgroundAdenoma detection rate (ADR) is a quality indicator for screening colonoscopy, but its calculation is time-consuming. Polyp detection rate (PDR) has been found to correlate with ADR; however, its use as a quality indicator has been criticized out of concern for endoscopists artificially inflating the PDR. We aim to evaluate whether active monitoring affects PDR.MethodsIn March 2015, 14 endoscopists were made aware that their personal PDRs would be tracked monthly as a quality improvement project. Endoscopists received a report of their individual monthly and cumulative PDR, departmental averages, and a benchmark PDR. Following the intervention, data were collected for consecutive patients undergoing average risk screening colonoscopy for six months. PDR, ADR, and adenoma to polyp detection ratio quotient (APDRQ) were compared to a six-month pre-intervention period.Results2203 patients were included in the study. There was no statistically significant difference in PDR when comparing pre- and post-intervention (44 vs. 45%, OR 1.04; 95% CI 0.77–1.36). No statistically significant difference in ADR was observed when comparing pre- and post-intervention (29 vs. 30%, OR 1.03; 95% CI 0.64–1.52). There was no statistically significant difference in APDRQ when comparing pre- and post-intervention (0.67 vs. 0.66, OR 0.99; 95% CI 0.69–1.33).ConclusionsMonthly report cards did not result in a change in PDR or APDRQ. In some environments, PDR can be used as a surrogate marker of ADR, despite endoscopist awareness that PDR is being measured.

Transocular ultrasound measurement of the optic nerve sheath diameter can identify elevated intracranial pressure in trauma patients

Background Prompt identification of elevated intracranial pressure (eICP) is life-saving. Physical exam alone often fails to identify this problem and invasive monitoring is not always utilized appropriately. A non-invasive, rapid, reliable technique to detect eICP is needed and optic nerve sheath diameter (ONSD) is known to be a valid surrogate. Historically, ONSD in patients with eICP above 20 mmHg varied from 0.5 to 0.7 cm. Receiver operator curves predict that diameters below 0.5 to 0.58 cm correspond to normal pressures. Interobserver variability is low (0.01--0.03 cm). The learning curve was 25 and 10 procedures for novice and experienced sonographers, respectively. We report our initial experience using ultrasound to measure ONSD during initial trauma evaluation. Study design Transocular ultrasound was used to determine ONSD after traumatic brain injury. Data were correlated with CT and exam findings or measured ICP. Measurements were performed by a single trauma surgeon. Time to perform the measurement was recorded. Relevant literature from the past 15 years was reviewed. Results Physical exam was not a reliable indicator of eICP. In patients with eICP requiring surgery, ONSD averaged 0.62 cm. Normal pressures were ensured if the ONDS was <0.50 cm (95% CI 0.469–0.540 cm, p < 0.001). Intraobserver variability was low (0.01--0.02 cm). Average time to perform the exam was less than 2 min. Conclusions Determining ONSD with ultrasound is easy to learn, precise, and inexpensive. An ONSD of less than 5 mm ensures no eICP. This procedure can be added to the evaluation of trauma patients with no delay in care. Future prospective studies may validate and incorporate this technique into the trauma surgeons armamentarium.