Assessing the reproducibility of American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) arthroplasty studies

Table 3 Summary of reproduced adjusted odds ratios (aORs). Specifically, reproduced aORs that changed in interpretation from the original study, changes in statistical significance in reproduced aORs, and changes in sample size between original and reproduced data are included

Study Number	aORs				Statistical Significance*		Comparison of Sample Size (N)
Study Number	Total aORs in Study	Reproduced aORs that changed in interpretation	Total aOR Change (%)	Average Magnitude Change of aOR (%)	Reproduced aORs that changed in significance	Significance Change (%)	Original Total (N)	Reproduced Total (N)	Difference in Sample Size	Difference in Sample Size (%)
1	15	4	26.67	9.40	5	33.33	78,191	83,736	5545	7.09
2	42	3	7.14	2.10	6	14.29	5068	5198	130	2.57
3	56	2	3.57	26.34	5	8.93	10,112	9788	324	3.20
4	40	5	12.50	17.01	5	12.50	169,406	169,406	0	0.00
5	35	4	11.43	39.49	1	2.86	120,742	167,402	46,660	38.64
6	3	0	0.00	17.30	2	66.67	8237	8790	553	6.71
7	3	0	0.00	13.67	1	33.33	8776	7945	831	9.47
8	3	0	0.00	47.00	2	66.67	2502	2476	26	1.04
9	24	0	0.00	6.79	0	0.00	210,075	210,064	11	0.01
10	7	2	28.57	4.99	5	71.43	67,897	35,413	32,484	47.84
11	40	14	35.00	5.35	4	10.00	2088	2208	120	5.75
Total	268	34	12.69%	17.22%	36	13.43%	683,094	702,426	19,332	2.83%

ISSN: 1749-799X