Extending fecal immunochemical tests for hemoglobin (FITs) to primary care patients with high-risk symptoms suggestive of colorectal cancer (CRC) could reduce colonoscopy waiting lists, enabling earlier treatment. Higher FIT thresholds could decrease referrals but increase missed disease compared with lower thresholds. We aimed to systematically review and synthesize test accuracy data across thresholds for use in a cost-effectiveness analysis.

Searches across ten sources were conducted (December 2022). Included were diagnostic accuracy studies of HM-JACKarc, OC-Sensor, FOB Gold, QuikRead go, NS-Prime, and four Immunodiagnostik (IDK) tests in patients presenting to, or referred from, primary care with symptoms suggestive of CRC using any reference standard. Risk of bias was assessed with QUADAS-2. Syntheses of sensitivity and specificity at all reported thresholds were planned for each test to provide summary estimates at all possible thresholds within the observed range. Sensitivity analyses investigating population type and reference standard, and subgroup analyses by patient characteristics (e.g., anemia, age, sex, ethnicity) were conducted.

HM-JACKarc (n=16 studies) sensitivity ranged from 95.9 percent (95 percent credible interval [95% CrI]: 92.7, 97.9) to 46.3 percent (95% CrI: 37.4, 54.9) and specificity from 65.1 percent (95% CrI: 55.6, 74.8) to 97.7 percent (95% CrI: 94.7, 99.2) (thresholds 2 and 400 μg hemoglobin/g feces [μg/g], respectively). OC-Sensor (n=11) sensitivity ranged from 94.2 percent (95% CrI: 91.2, 96.7) to 54.2 percent (95% CrI: 48.4, 60.2) and specificity from 62.7 percent (95% CrI: 47.4, 77.2) to 97.3 percent (95% CrI: 92.9, 99.3) (thresholds 4 and 200 μg/g, respectively). FOB Gold (n=3) sensitivity ranged from 91.4 percent (95% CrI: 71.6, 99.6) to 73.9 percent (95% CrI: 53.8, 91.2) and specificity from 78.1 percent (95% CrI: 70.0, 86.0) to 96.4 percent (95% CrI: 92.6, 98.9) (thresholds 2 and 150 μg/g, respectively). There were limited or no data on the other tests.

Sensitivity and specificity were synthesized for three tests only, since data for the remaining tests were extremely limited or absent. Even at the lowest threshold, none of the tests had perfect sensitivity. Future studies should further investigate comparative accuracy and the impact of patient characteristics, patient recruitment criteria, and the reference standard on estimates of diagnostic test accuracy.

Approximately 42,000 new cases of colorectal cancer (CRC) are diagnosed annually in the United Kingdom with 16,800 deaths. Evidence suggests that quantitative fecal immunochemical tests (FIT) are a good predictor of CRC risk in symptomatic patients presenting to primary care. We aimed to assess the cost-effectiveness of FIT in this setting, considering capacity constraints and waiting times for subsequent colonoscopy.

We compared two diagnostic FIT strategies, at various thresholds, in the model: (i) FIT for all patients and (ii) current practice where only low-risk patients received FIT. Patients with positive FIT scores and high-risk patients in current practice received colonoscopy. Diagnostic accuracy evidence from published literature, standard UK cost sources, and other sources were used to estimate health outcomes and costs. Waiting times before colonoscopy were assumed proportional to the numbers referred, with the impact of delayed colonoscopy taken from published models. Savings per quality-adjusted life years (QALYs) lost and incremental net monetary benefit (INMB) were used. Uncertainty was evaluated.

Model results suggested that, compared to current practice, FIT generated a positive INMB for the majority of thresholds assessed (GBP200 [USD254] to GBP350 [USD445] per patient at a willingness to pay of GBP20,000 [USD25,474] per QALY gained). A reduction in the number of patients sent to colonoscopy led to cost savings. However, these thresholds were associated with slight QALY losses due to a small proportion of false negative results associated with significantly delayed diagnosis, which outweighed the benefits associated with quicker times to colonoscopy for those with positive FIT results. Savings of over GBP100,000 (USD127,374) per QALY lost were generated. Conclusions were robust to the sensitivity analyses undertaken.

With capacity constraints explicitly represented in the economic modeling, offering FIT to all patients presenting to primary care with symptoms suggestive of CRC was cost effective when compared to current practice. However, the optimal threshold could not be robustly determined due to limited diagnostic accuracy data, parameter uncertainty, and limitations in the model structure; additional primary research could reduce uncertainty.

Diagnostic criteria for major depressive disorder allow for heterogeneous symptom profiles but genetic analysis of major depressive symptoms has the potential to identify clinical and etiological subtypes. There are several challenges to integrating symptom data from genetically informative cohorts, such as sample size differences between clinical and community cohorts and various patterns of missing data.

We conducted genome-wide association studies of major depressive symptoms in three cohorts that were enriched for participants with a diagnosis of depression (Psychiatric Genomics Consortium, Australian Genetics of Depression Study, Generation Scotland) and three community cohorts who were not recruited on the basis of diagnosis (Avon Longitudinal Study of Parents and Children, Estonian Biobank, and UK Biobank). We fit a series of confirmatory factor models with factors that accounted for how symptom data was sampled and then compared alternative models with different symptom factors.

The best fitting model had a distinct factor for Appetite/Weight symptoms and an additional measurement factor that accounted for the skip-structure in community cohorts (use of Depression and Anhedonia as gating symptoms).

The results show the importance of assessing the directionality of symptoms (such as hypersomnia versus insomnia) and of accounting for study and measurement design when meta-analyzing genetic association data.