{"args":{"level":0.95,"unit":"log relative risk"},"args_from_frame":{"estimates":{"dataset":"bcg","fields":{"estimate":"yi","id":"trial","se":"sei","study_id":"trial"}}},"checks":[{"abs_tol":1.7874208953540318e-07,"bind":"any","column":null,"failure_class":null,"field":"value","interval":false,"key":"pooled","kind":"figure_close","min_lre":null,"note":"software reference","open_interval":false,"rel_tol":1e-06,"required":true,"source_id":"colditz_1994.trials_rr.software","table":null,"tool":"evidence.pool_dl","value":-0.7141172220732273,"weight":1.0,"where":null},{"abs_tol":1.7874208953540318e-07,"bind":"any","column":null,"failure_class":null,"field":"se","interval":false,"key":"pooled","kind":"figure_close","min_lre":null,"note":"software reference, SE","open_interval":false,"rel_tol":1e-06,"required":true,"source_id":"colditz_1994.trials_rr.software","table":null,"tool":"evidence.pool_dl","value":0.1787420895354032,"weight":1.0,"where":null}],"datasets":{"bcg":{"columns":null,"description":"The trials of BCG vaccine against tuberculosis analysed by Colditz et al. (1994), as metadat's dat.bcg holds them: one row per trial. Columns: trial, the trial number; author and year of the report; tpos and tneg, the vaccinated people who did and did not get tuberculosis; cpos and cneg, the same for the unvaccinated; ablat, the trial's absolute latitude; alloc, how treatment was allocated. Built columns: n_t and n_c, the arm sizes; risk_t and risk_c, the arms' tuberculosis risks; rr, their ratio; yi, the log relative risk; vi, its large-sample variance (tneg/(tpos\u00b7n_t) + cneg/(cpos\u00b7n_c)); sei, its standard error.","file_sha256":"8508024604252ddecbbba88a5ebbbf76ddec7d32f9e5df56521f602ce5398fdb","format":"rdata","frame_sha256":null,"generator":null,"kind":"file","licence":"metadat R package (GPL-2 or later); fetched, not redistributed.","member":null,"object":"dat.bcg","params":{},"path":null,"read_kw":{},"records":null,"role":"agent","sha256":"8508024604252ddecbbba88a5ebbbf76ddec7d32f9e5df56521f602ce5398fdb","transforms":[{"args":{"columns":["trial"],"to":"str"},"op":"cast"},{"args":{"args":{"columns":["tpos","tneg"]},"name":"n_t","op":"sum"},"op":"derive"},{"args":{"args":{"columns":["cpos","cneg"]},"name":"n_c","op":"sum"},"op":"derive"},{"args":{"args":{"left":"tpos","right":"n_t"},"name":"risk_t","op":"ratio"},"op":"derive"},{"args":{"args":{"left":"cpos","right":"n_c"},"name":"risk_c","op":"ratio"},"op":"derive"},{"args":{"args":{"left":"risk_t","right":"risk_c"},"name":"rr","op":"ratio"},"op":"derive"},{"args":{"args":{"column":"rr"},"name":"yi","op":"log"},"op":"derive"},{"args":{"args":{"columns":["tpos","n_t"]},"name":"tpos_n_t","op":"product"},"op":"derive"},{"args":{"args":{"columns":["cpos","n_c"]},"name":"cpos_n_c","op":"product"},"op":"derive"},{"args":{"args":{"left":"tneg","right":"tpos_n_t"},"name":"v_t","op":"ratio"},"op":"derive"},{"args":{"args":{"left":"cneg","right":"cpos_n_c"},"name":"v_c","op":"ratio"},"op":"derive"},{"args":{"args":{"columns":["v_t","v_c"]},"name":"vi","op":"sum"},"op":"derive"},{"args":{"args":{"column":"vi"},"name":"log_vi","op":"log"},"op":"derive"},{"args":{"args":{"column":"log_vi","factor":0.5},"name":"half_log_vi","op":"scale"},"op":"derive"},{"args":{"args":{"column":"half_log_vi"},"name":"sei","op":"exp"},"op":"derive"}],"url":"https://raw.githubusercontent.com/wviechtb/metadat/b19c23b4c09e16d0c14a31af94fca190c7dd0357/data/dat.bcg.rda"}},"difficulty":"E","id":"paper_repro.colditz_1994.trials_rr.tool","level":"T1","meta":{"acceptable_tools":["evidence.pool_dl"],"agent_excluded":"The agent cannot pass per-study estimates to evidence.pool_dl without typing numbers: no tool builds pool_dl's estimates from a frame, and the model never types a number (a coverage gap).","cluster":"colditz_1994","contamination":{"benchmark_first_public":"1994-03","gold_in_public_repo":true,"model_reported_cutoff":null,"url":"https://pubmed.ncbi.nlm.nih.gov/8309034/"},"expected_tool":"evidence.pool_dl","family":"meta","kind":"tool","known_deviation":{"evidence":"Recorded before any software-reference run or dry run, from the abstract and the data's provenance: the abstract pools fourteen prospective trials, the public trial-level data hold 13, and the abstract does not say which random-effects estimator it used.","fields":["se","value"],"found_in_run":false,"kind":"data_version"},"locator":"Abstract, Data Synthesis: random-effects relative risk of TB in the prospective trials","n_published":14,"paper":"colditz_1994","printed_not_compared":"The abstract prints the pooled relative risk as a ratio with a 95% interval (0.49, 0.34 to 0.70) on 14 trials and no standard error; the tool pools log relative risks of the 13 public trials. A ratio is not the tool's figure, so the printed cell is cited and recorded, not compared (known deviation data_version).","se_policy":{"basis":"The abstract states a random-effects model and prints an interval, not an SE; the estimator is not named.","convention":"DerSimonian-Laird random-effects standard error of the pooled mean","never_required":false,"stated_in_question":false},"target":"trials_rr","trap_tools":[]},"question":"","requires":[],"runner":null,"sources":["colditz_1994.trials_rr.paper","colditz_1994.trials_rr.software"],"split":"public","suite":"paper_repro","tags":["paper_repro","colditz_1994","meta"],"target":"tool","tool":"evidence.pool_dl","truth":"Colditz, G. A., Brewer, T. F., Berkey, C. S., Wilson, M. E., Burdick, E., Fineberg, H. V. and Mosteller, F. (1994). Efficacy of BCG Vaccine in the Prevention of Tuberculosis: Meta-analysis of the Published Literature. JAMA 271(9): 698-702. Abstract, Data Synthesis: random-effects relative risk of TB in the prospective trials (source colditz_1994.trials_rr.paper); software reference colditz_1994.trials_rr.software: statsmodels 0.15.0: the reference specification of colditz_1994 trials_rr (evidence.pool_dl, figure 'pooled') on an independent code path"}