STUDY | STUDY POPULATION | STUDY GROUPS | DESIGN | RESULTS | COMMENTS |
---|---|---|---|---|---|
Fries and Britton (1973)27 | Rheumatoid arthritis (N = 27) (mean age 56.3 y) | Fenoprofen (1.4–2.4 g/d) for 6 wk High-dose ASA (4–6 g/d) for 6 wk Placebo for 3 wk | RCT, double-blind crossover with temporal controls | Placebo: 15.4% positive FOBT results ASA: 17.1% positive results Fenoprofen: 20% positive results No significant difference between groups | Unspecified guaiac-based FOBT; true positives not determined |
Greenberg et al (1999)28 | Healthy house officer volunteers (mean age 29.8 y) | ASA (30, 81, or 325 mg/d) for 30 d (n = 10 in each group) Placebo for 30 d (n = 10) | RCT, double blind | 0 positive results using Hemoccult II or Hemoccult SENSA, in all groups | |
Brooks et al (1970a)29 | Rheumatoid arthritis (mean age 51.5 y) | Ibuprofen (600 mg/d) for 4 wk (n = 41) ASA (3.6 g/d) for 4 wk (n = 45) U-24568 (600 mg/d) for 4 wk (n = 41) | Randomized, double blind | Total positive FOBT results not clarified No significant differences between groups | No control group; unspecified guaiac-based FOBT |
Brooks et al (1970b)29 | Rheumatoid arthritis (mean age 53.1 y) | Ibuprofen (900 mg/d) for 4 wk (n = 60) ASA (3.6 g/d) for 4 wk (n = 62) | Randomized, double blind | Total positive FOBT results not clarified No significant differences between groups | No control group; unspecified guaiac-based FOBT |
Zuin et al (2000)30 | Healthy volunteers (mean age 27.7 y) | Ibuprofen tablet (800 mg/d) for 7 d (n = 18) Ibuprofen fast-melting tablets (800 mg/d) for 7 d (n = 18) | Two-sequence crossover | 0 positive FOBT results | No control group; unspecified FOBT |
Hurlen et al (2006)31 | AMI survivors (mean age 60.9 y) | ASA (160 mg/d) for 3 mo (n = 94) Warfarin (INR 2.8–4.2) for 3 mo (n = 84) ASA (75 mg/d) plus warfarin (INR 2.0–2.5) for 3 mo (n = 89) | Randomized | ASA: 8.5% positive hemo FEC results Warfarin: 7.1% positive results ASA with warfarin: 5.6% positive results No significant differences between groups | No control group; 14 of 19 positive results retested (3 remained positive, 2 with diverticulosis and 1 with normal findings on colonoscopy) |
Bahrt et al (1984)32 | Rheumatoid disease clinic patients (age not reported) | Unspecified NSAID, salicylate, or steroid use (n = 145) | Cross-sectional | Anti-inflammatory: 5.5% positive results using Hemoccult | No control group; all positive results evaluated (2 colonic neoplasms [PPV 25%]) |
Norfleet (1983)33 | Healthy volunteers (N = 27) | Control period for 3 d ASA (1300 mg/d) for 7 d | Temporally controlled trial | 0 positive Hemoccult II test results during both test periods | |
Greenberg et al (1996)34 | Medical, cardiac, and anticoagulation clinic patients (mean age approximately 60 y) | Control for 1 wk, then ASA 325 mg/d for 8 wk (n = 25) ASA 325 mg/d for 1 wk, then ASA 81 mg/d for 8 wk (n = 46) ASA 81 mg/d for 1 wk, then ASA 325 mg/d for 8 wk (n = 4) Warfarin (unspecified INR) for 4–6 wk (n = 25) | Cross-sectional with crossover | Control week: 0 positive results using Hemoccult II ASA 81 mg/d: 14% positive results ASA 325 mg/d: 4% positive results Warfarin: 12% positive results No significant differences between groups | Blood loss quantified with HemoQuant; no difference between groups |
Niv (1987)35 | Israeli screening population taking NSAIDs (aged > 40 y) | Self-reported discontinuing NSAIDs during FOBT (n = 1771) Self-reported NSAID use during FOBT (n = 26) | Cohort | NSAID users: 27% positive results using Hemoccult II Nonusers: 4% positive results (P < .01) | Relied on self-reporting of compliance with FOBT instructions |
Pye et al (1987)36 | Asymptomatic subjects with positive screening results on Hemoccult or Feca-EIA (aged 50–74 y) | NSAID use according to self-report (n = 50) No NSAID use according to self-report (n = 405) | Cross-sectional | 4.2% total positive Hemoccult or Feca-EIA results from screening population of 10 931 All positive results evaluated with colonoscopy NSAID users: PPV 20% for neoplasia Nonusers: PPV 32% No significant difference (P = .1) | Hemoccult and Feca-EIA (immunologic test) results not reported separately |
Kahi and Imperiale (2004)37 | VA patients referred for colonoscopy with positive Hemoccult II FOBT results (N = 193) (mean age 66 y) | ASA or NSAID use according to self-report and record review (n = 135) No ASA or NSAID use according to self-report and record review (n = 58) | Cross-sectional | ASA or NSAID users: PPV 21% for “abnormality” Nonusers: PPV 19% No significant differences between groups | No correlation between ASA dose and colonoscopic pathology; not a screening population; large polyps and CRC were not reported separately from other colonic pathology |
Clarke et al (2006)38 | Scottish patients referred for colonoscopy with positive Hema Screen (guaiac) FOBT results (aged 50–69 y) | Self reported ASA, NSAID, or anticoagulant use (n = 301) No ASA, NSAID, or anticoagulant use (n = 308) | Cross-sectional | ASA, NSAID, or anticoagulant users: PPV 47.5% for colorectal neoplasia Nonusers: PPV 56.5% (P = .012) | PPV for CRC with no significant differences between groups (P = .7); anticoagulants made up only 7.7% of prescriptions |
Sawhney et al (2010)39 | VA patients referred for colonoscopy with positive Hemoccult II results (mean age approximately 68 y) | No medications according to review of pharmacy profile (n = 518) ASA (n = 264) NSAIDs (n = 218) Warfarin (n = 85) Clopidogrel (n = 41) | Cross-sectional | Controls: PPV 30.5% for advanced neoplasia ASA users: PPV 20.5% (P < .01) NSAID users: PPV 19.7% (P < .01) Warfarin users: PPV 20% (P = .05) Clopidogrel users: PPV 7.3% (P < .01) | Not a screening population; included positive FOBT results for evaluation of symptoms |
Doran and Hardcastle (1982)40 | CRC patients and age-matched controls | Temporal control without ASA, then ASA (600 mg/d) for 3 d (n = 50) Age-matched subjects with temporal control, then ASA (600 mg/d) for 3 d (n = 50) | Temporally controlled trial with control cohort component | CRC patients during temporal control: 70% positive Hemoccult II results CRC patients taking ASA: 82% positive results (P = .07) Age-matched cohort: single subject with positive Hemoccult II results before and while taking ASA | 51Cr-RBC labeling in CRC patients (n = 25) showed no correlation between blood volume and ASA use or tumour location |
Kewenter et al (1984)41 | Patients taking dicumarol (N = 849) (mean age 67 y) | Dicumarol, unspecified dose | Cross-sectional | 15% positive results using Hemoccult II PPV 19% for CRC and adenomas among 79 subjects evaluated further (57 of 67 with ≥ 3 of 6 positive samples; 22 of 61 with < 3 of 6 positive samples) | No correlation between positive Hemoccult result and anticoagulation index |
Jaffin et al (1987)42 | Patients taking warfarin, heparin, or both (n = 256) and age- and diagnosis-matched inpatient controls (n = 164) (mean age 64 y) | Anticoagulant users (Hemoccult completed in n = 175) Age- and diagnosis-matched controls (Hemoccult completed in n = 74) | Controlled cross-sectional | Anticoagulant users: 12% positive results using Hemoccult Controls: 3% positive results (P < .01) | No correlation with anticoagulation index; evaluation of 16 of 21 positive test results among anticoagulant users (2 with CRC [PPV 12.5%]); no evaluation of positive test results in control group |
Kershenbaum et al (2010)43 | Israeli CRC screening program participants (aged 50–74 y) | Warfarin treatment (n = 1356 FOBTs) No antithrombotic or anticoagulant use (n = 64 088 FOBTs) | Cohort | Warfarin users: 7.7% positive results using Hemoccult SENSA Nonusers: 3.6% positive results (P < .01) Warfarin users: PPV 19.8% for CRC or clinically significant adenoma in 81.9% of positive FOBT results Nonusers: PPV 27.7% in 74.7% of positive FOBT results No significant difference in PPV between groups | |
Bini et al (2005)44 | Patients referred for colonoscopy to evaluate positive Hemoccult II results (mean age 72 y) | Warfarin users (n = 210) Age- and sex-matched controls (n = 210) | Controlled, cross-sectional | Warfarin users: PPV 27.2% for CRC or adenomas Nonusers: PPV 24.3% for CRC or adenomas No significant difference in PPV between groups (P = .58) | Not a screening population; included positive FOBT results obtained for evaluation of symptoms |
Iles-Shih et al (2010)45 | Patients referred for colonoscopy to evaluate positive screening results using Hemoccult II (aged < 50 to > 80y) | Warfarin users (n = 372) Warfarin nonusers (n = 9265) | Cross-sectional | Warfarin users: PPV 16.1% for large polyps or tumours Nonusers: PPV 11.4% No significant difference in PPV between groups after adjusting for age and sex | INRs not reported |
51Cr-RBC—chromium-51 red blood cell, AMI—acute myocardial infarction, ASA—acetylsalicylic acid, CRC—colorectal cancer, FOBT—fecal occult blood testing, INR—international normalized ratio, NSAID—nonsteroidal anti-inflammatory drug, PPV—positive predictive value, RCT—randomized controlled trial, VA—Veterans Affairs.