A71 POLYP TO ADENOMA CONVERSION FACTOR AS A SURROGATE FOR ADENOMA DETECTION RATE-– FINDINGS FROM THE SOUTHWEST ONTARIO COLONOSCOPY COHORT

Background

The adenoma detection rate (ADR) is one of the main quality indicators of a colonoscopy but requires combining endoscopic and histologic data. However, the polyp detection rate (PDR) requires only endoscopic assessment and has been proposed as a proxy measure for the ADR.

Aims

To calculate a conversion factor for PDR to ADR, for use as a future surrogate of ADR when only PDR is available.

Methods

The Southwest Ontario Colonoscopy cohort consists of all outpatient colonoscopies performed across 20 hospitals in Southwestern Ontario between April 2017 and February 2018. Data was collected prospectively through a mandatory quality assurance form that was completed after each procedure and pathology reports were manually reviewed. Endoscopies with associated histologic findings were included.

The PDR and true ADR were calculated for each physician. A weighted polyp to adenoma detection rate quotient (APDRQ) was calculated, weighting each physician’s APDRQ by the number of procedures performed. The APDRQ was determined for all outpatient procedures and specifically for screening/surveillance indications.

Results

During the study period, 57 endoscopists performed 31,721 colonoscopies. The overall PDR was 41.1% and the ADR was 26.5%. The weighted ADPDRQ was 0.638 (95% CI: 0.600, 0.675). When limited to screening/surveillance colonoscopies, the weighted ADPDR was 0.616 (95% CI: 0.564, 0.669). To better understand the influence of endoscopists with low ADR: PDR, we excluded those with ratio below (<2 standard errors) the average, which resulted in greater ADR: PDR for all colonoscopies 0.695 (95% CI: 0.679, 0.711) and for screening/surveillance colonoscopies and 0.692 (95% CI: 0.677, 0.707).

Conclusions

In this large, population-based, cohort study, we calculated the ADR; PDR ratio. We propose this may be used in future studies to infer ADR when only PDR is available.

Scatter plot of correlation between ADR and PDR, by physician. The dashed line indicates the line for which ADR=PDR, the maximum value the ADR can take for a given PDR. The marker size is proportional to the number of colonoscopies performed.

Funding Agencies

None

A113 ANNUAL COLONOSCOPY VOLUME IS NOT PREDICTIVE OF COLONOSCOPY QUALITY - FINDINGS FROM THE SOUTHWEST ONTARIO COLONOSCOPY COHORT

Background

Performing a minimum number of colonoscopies annually has been proposed by some jurisdictions as a requirement for maintaining privileges. However, this practice is supported by limited evidence.

Aims

The objective of this study was to determine if annual colonoscopy volume was associated with colonoscopy quality metrics.

Methods

A population-based study was performed using the Southwest Ontario Colonoscopy cohort, which consists of all adult patients who underwent colonoscopy between April 2017 and Oct 2018 at 21 academic and community hospitals within the health region. Data were collected through a mandatory quality assurance form completed after each procedure and pathology reports were manually reviewed. Physician annualized colonoscopy volumes were compared by correlation analysis to each quality-related outcome, by means of the area under the receiver operating characteristics curve (AUROC), and logistic regression. The prognostic value of colonoscopy volume was also adjusted for case-mix and potential confounders in separate regression analyses for each outcome. The primary outcome was ADR. Secondary outcomes were polyp detection rate (PDR), sessile serrated polyp detection rate (SSPDR), and cecal intubation.

Results

A total of 47,195 colonoscopies were performed by 75 physicians (37.5% by gastroenterologists, 60% by general surgeons, 2.5% others). There were no clear relationships between annual colonoscopy volumes and study outcomes. Colonoscopy volume was not associated with ADR (OR 1.03, 95% CI 0.96–1.10, p=0.48) and corresponded to an AUROC not significantly different from the null (AUROC 0.52, 95% CI 0.43–0.61, p=0.65). Multi-variable regression adjusting for case-mix also demonstrated no predictive value of annual colonoscopy volume for the primary outcome (OR 1.03, 95% CI 0.94–1.12, p=0.55). Similarly, analyses of secondary outcomes failed to find an association between colonoscopy volume and PDR, SSPDR, or cecal intubation (Table 1).

Conclusions

Annual colonoscopy volumes do not predict ADR, PDR, SSPDR, or cecal intubation rate.

Results of unconditional and conditional approaches for examining the predictive value of annual colonoscopy volume for quality related outcomes.

Funding Agencies

None

A117 PROPOFOL SEDATION DOES NOT IMPROVE MEASURES OF COLONOSCOPY QUALITY – FINDINGS FROM THE SOUTHWEST ONTARIO COLONOSCOPY COHORT

Background

The use of propofol during colonoscopy has gained increased popularity due to deeper anesthesia compared to conscious sedation. Prior studies examining the use of propofol sedation during colonoscopy have primarily focused on anesthesia outcomes. Whether propofol sedation is associated with improvements in colonoscopy outcomes is uncertain.

Aims

The primary outcome was adenoma detection rate (ADR). Secondary outcomes were the detection of any adenoma (conventional adenoma, sessile serrated polyp, and traditional serrated adenoma), sessile serrated polyp detection rate, polyp detection rate, cecal intubation rate, and perforation rate.

Methods

The Southwest Ontario Colonoscopy cohort consists of all patients who underwent colonoscopy between April 2017 and Oct 2018 at 21 hospitals serving a large geographic area in Southwest Ontario. Procedures performed in patients less than 18 years of age or by endoscopist who perform &lt;50 colonoscopies/year were excluded. Data were collected through a mandatory quality assurance form that was completed by the endoscopist after each procedure. Pathology reports were manually reviewed.

Results

A total of 46,634 colonoscopies were performed by 75 physicians (37.5% by gastroenterologists, 60% by general surgeons, 2.5% others) of which 16,408 (35.2%) received propofol and 30,226 (64.8%) received conscious sedation (e.g. combination of a benzodiazepine and a narcotic). Patients who received propofol were likely to have a screening indication (49.2% vs 45.5%, p&lt;0.0001), not have a trainee endoscopist present and be performed at a non-academic centre (32.2% vs 44.6%, p&lt;0.0001). Compared to conscious sedation, use of propofol was associated with a lower ADR (24.6% vs. 27.0%, p&lt;0.0001) and detection of any adenoma (27.7% vs. 29.8%, p&lt;0.0001); no difference was observed in the detection ofsessile serrated polyps (5.0% vs. 4.7%, p=0.26), polyp detection rate (41.2% vs 41.2%, p=0.978), cecal intubation rate (97.1% vs. 96.8%, p=0.15) or perforation rate (0.04% vs. 0.06%,p=0.45). On multi-variable analysis, the use of propofol was not significantly associated with any improvement in ADR (RR=0.90, 95% CI 0.74–1.10, p=0.30), detection of any adenoma (RR=0.93, 95% CI 0.75–1.14, p=0.47), sessile serrated polyp detection rate (RR=1.20, 95%CI 0.90–1.60, p=0.22), polyp detection rate (RR=1.00, 95% CI 0.90–1.11, p=0.99), or cecal intubation rate (RR=1.00, 95%CI 0.80–1.26, p=0.99).

Conclusions

The use of propofol sedation does not improve colonoscopy quality metrics.

Funding Agencies

None

A115 A SYSTEMATIC REVIEW AND META-ANALYSIS OF LOWER GASTROINTESTINAL BLEEDING RISK SCORES TO PREDICT ADVERSE OUTCOMES

Background

Acute lower gastrointestinal bleeding (LGIB) is a common reason for emergency hospitalization. In most patients, bleeding resolves spontaneously but some cases can be fatal. Risk prediction scores can be useful in risk stratifying patients with LGIB at the time of presentation although the most discriminative LGIB risk score is unknown.

Aims

To perform a systematic review and meta-analysis comparing LGIB risk prediction scores.

Methods

Following the PRISMA statement, a systematic search for relevant publications after 1990 was conducted in Ovid Medline, EMBASE, Web of Science and CENTRAL electronic databases. We also searched published conference abstracts over the past 5 years. Studies with a primary aim of deriving or validating a LGIB risk score were included. Title and abstracts were reviewed by two independent reviewers followed by full text review and data extraction by both reviewers. Diagnostic classification data for combinations of risk score and clinical outcome were meta-analyzed using a hierarchical summary receiver operator characteristic curve (ROC) model, allowing for random-effects by study, and fixed-effect of the risk score thresholds to influence both sensitivity and specificity. Area under the summary ROC were estimated from model parameters for the pre-specified LGIB risk score thresholds-of-interest.

Results

Our search identified 2,331 citations for review, of which 100 remained after the title and abstract screen, and 18 ultimately met criteria for inclusion in the meta-analysis after full text review. From these, we identified 21 risk prediction scores for LGIB, although only four had sufficient number of papers to meta-analyze (Oakland, Strate, NOBLADS, and BLEED score). For the outcome safe discharge from hospital, the Oakland score had an area under the receiver operating characteristics curve (AUROC) of 85.5% (95% CI: 82.1%, 88.3%). For the outcome major bleeding, the Oakland score had an AUROC of 78.9% (95% CI: 75.1%, 82.2%); the Strate score had an AUROC of 74.4% (95% CI: 70.4%, 78.0%); the NOBLADS score had an AUROC of 60.3% (95% CI: 55.9%, 64.5%); and the BLEED score had an AUROC of 65.6% (95% CI: 61.4%, 69.7%). For the outcome, need for hemostasis, the Oakland had an AUROC of 99.0% (95% CI: 97.7%, 99.6%); the Strate score had an AUROC of 82.1% (95% CI: 78.5%, 85.2%); the NOBLADS score had an AUROC of 23.9% (95% CI: 20.3%, 27.8%). For the outcome, need for transfusion, the Oakland score had an AUROC of 99.0% (95% CI: 97.7%, 99.6%); the NOBLADS score had an AUROC of 87.7% (95% CI: 84.5%, 90.3%).

Conclusions

The Oakland score was the most discriminative risk prediction model for safe discharge from hospital, major bleeding, need for hemostasis, and need for transfusion.

Funding Agencies

None

A120 NO EVIDENCE OF A FRIDAY EFFECT ON COLONOSCOPY QUALITY OUTCOMES

Background

Colonoscopy quality may be influenced by operator fatigue. Prior studies have shown lower adenoma detection rates for procedures performed at the end of the day. However, it is unknown if colonoscopy quality is impaired at the end of the work week.

Aims

We investigated whether colonoscopy quality-related metrics differ at the end of the work week using the South West Ontario Colonoscopy Cohort.

Methods

Between April 2017 to February 2018, 45,510 consecutive colonoscopies from 20 academic and community hospitals in our health region were captured to form the cohort. In Canada, outpatient endoscopies are generally performed between Monday to Friday, taking Friday, or the last business day, as the last day of the work week compared to the rest of the work week. When a statutory holiday occurred on a Friday, Thursday was designated the last day of the work week. The primary outcome was adenoma detection rate (ADR), and secondary outcomes were sessile serrated polyp detection rate (ssPDR), polyp detection rate (PDR), and failed cecal intubation. Outcomes were presented as unadjusted and adjusted risk ratios derived from modified Poisson regression and adjusting for physician-level clustering, and characteristics of the patient (age, sex, severity), procedure (hospital setting, trainee presence, indication, sedation, bowel preparation quality) and physician (experience and specialty).

Results

During the observation period, 9,132 colonoscopies were performed on the last day of the work week compared to 36,378 procedures during the rest of the work week. No significant difference was observed for ADR (26.4% vs. 26.6%, p=0.75), ssPDR (4.5% vs. 5.0%, p=0.12), PDR (44.1% vs. 43.1%, p=0.081), or failed cecal intubation (2.8% vs. 2.9%, p=0.51) for colonoscopies performed on the last day of the work week compared to the rest of the week, respectively. After adjusting for potential confounders, there were no significant differences in the ADR (RR 1.01, 95% CI [0.88, 1.15], p=0.94), ssPDR (RR 0.90, 95% CI [0.70, 1.14], p=0.38), PDR (RR 1.00, 95% CI [0.92, 1.09], p=0.94), or failed cecal intubation (RR 0.92, 95% CI [0.72, 1.18], p=0.51) for colonoscopies performed on the last day of the work week compared to the rest of week, respectively.

Conclusions

Colonoscopy quality metrics, including ADR, ssPDR, PDR, and failed cecal intubation are not significantly different at the end of the week.

Funding Agencies

None

Systematic review with meta-analysis: prevalence, risk factors and costs of aminosalicylate use in Crohn's disease

BACKGROUND

Aminosalicylates are the most frequently prescribed drugs for patients with Crohn's disease (CD), yet evidence to support their efficacy as induction or maintenance therapy is controversial.

AIMS

To quantify aminosalicylate use in CD clinical trials, identify factors associated with use and estimate direct annual treatment costs of therapy.

METHODS

MEDLINE, Embase and CENTRAL were searched to April 2017 for placebo-controlled trials in adults with CD treated with corticosteroids, immunosuppressants or biologics. The proportion of patients co-prescribed aminosalicylates in placebo arms was pooled using a random-effects model. Meta-regression was used to identify factors associated with aminosalicylate use. Annual treatment costs were estimated using the 2016 Ontario Drug Benefit Program.

RESULTS

Forty-two induction and 10 maintenance trials were included. The pooled proportion of patients co-prescribed aminosalicylates was 44% [95% CI: 39%-49%] in induction trials and 49% [95% CI: 35%-64%] in maintenance trials. There was substantial to considerable heterogeneity (I²  = 86.0%, 91.8% for induction and maintenance trials, respectively). In multivariable meta-regression, aminosalicylate use has decreased over time in induction trials (OR 0.50 [95% CI: 0.34-0.74] per 10-year increment). While a decline has been seen over time, 35% of CD patients were still using aminosalicylates in contemporary trials from the last 5 years. The estimated annual cost for the lowest price mesalazine (mesalamine) formulation is approximately $32 million for the Canadian CD population.

CONCLUSIONS

Over one-third of CD patients entering clinical trials are still co-prescribed aminosalicylates. A definitive trial is needed to inform the conventional practice of using aminosalicylates as CD maintenance therapy.

Systematic review with meta-analysis: endoscopic and histologic placebo rates in induction and maintenance trials of ulcerative colitis

BACKGROUND

Regulatory requirements for claims of mucosal healing in ulcerative colitis (UC) will require demonstration of both endoscopic and histologic healing. Quantifying these rates is essential for future drug development.

AIMS

To meta-analyse endoscopic and histologic placebo response and remission rates in UC randomised controlled trials (RCTs) and identify factors influencing these rates.

METHODS

MEDLINE, EMBASE and the Cochrane Library were searched from inception to March 2017 for placebo-controlled trials of pharmacological interventions for UC. Endoscopic and histologic placebo rates were pooled by random effects. Mixed effects univariable and multivariable meta-regression was used to evaluate the influence of patient, intervention and trial-related study-level covariates on these rates.

RESULTS

Fifty-six induction (placebo n = 4171) and 8 maintenance trials (placebo n = 1011) were included. Pooled placebo endoscopic remission and response rates for induction trials were 23% [95 confidence interval (CI) 19-28%] and 35% [95% CI 27-42%] respectively, and 20% [95% CI 16-24%] for maintenance of remission. The pooled histologic placebo remission rate was 14% [95% CI 8-22%] for induction trials. High heterogeneity was observed for all outcomes (I² 56.2%-88.3%). On multivariable meta-regression, central endoscopy reading was associated with significantly lower endoscopic placebo remission rates (16% vs 25%; OR = 0.52, [95% CI 0.29-0.92], P = 0.03). On univariable meta-regression, higher histologic placebo remission was associated with concomitant corticosteroids (OR = 1.17 [95% CI 1.08-1.26], P < 0.0001, per 10% increase in corticosteroid use).

CONCLUSIONS

Placebo endoscopic and histologic rates range from 14% to 35% in UC RCTs but are highly heterogeneous. Outcome standardisation may reduce heterogeneity and is needed in this field.

Study on GRIA2, GRIA3 and GRIA4 genes highlights a positive association between schizophrenia and GRIA3 in female patients

Impairment of glutamatergic neurotransmission is one of the major hypotheses proposed to explain the neurobiology of schizophrenia. Therefore, the genes involved in the glutamate neurotransmitter system could be considered potential candidate genes for schizophrenia susceptibility. A systematic study on alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptor genes has been carried out and the results obtained from the analysis on GRIA2, GRIA3 and GRIA4 are reported. No evidence of association with schizophrenia was found for the GRIA2 and GRIA4 genes; strong evidence of association with schizophrenia was found for GRIA3. This X-linked gene showed a different behavior in the two genders; a positive association with schizophrenia was observed among females but not in males. Female carriers of rs1034428 A allele were found to have a 2.19-fold higher risk of developing schizophrenia compared to non-carriers and 3.28-fold higher risk for developing a non-paranoid phenotype. The analysis at the haplotype level showed that susceptibility to schizophrenia was associated with the specific haplotype rs989638-rs1034428-rs2227098 CAC (P = 0.0008). We conclude that, of the three AMPA genes analyzed here, only GRIA3 seems to be involved in the pathogenesis of schizophrenia, but only in females.