Table 1Evidence for Key Question 2: Studies of Performance Characteristics of the Oral Cancer Screening Examination

Author, year
Participants, behaviors, and setting*Criteria for positive screening testIdentity and training of screenersReference standardResultsCommentsQuality rating
Examination by health workers
Mehta et al, 198641

1,921 tobacco users age ≥35 years residing in Trivandrum District, India; proportion using alcohol NR

Subjects' homes
Presence of nodular leukoplakia, submucous fibrosis, ulcers, or growths suggestive of oral cancerBasic health workers (high school + 1 year) who received 2 or 5 days of training on how to perform oral exams and classify lesionsDentist's exam 6 months after screening examSensitivity=59% (16/27)
Specificity=98% (1,859/1,894)
PPV=31% (16/51)
NPV=99% (1,859/1,870)
Selected for reference exam from 39,331 who were screened. Unclear if dentists were aware of screening results when performing their exams. Study conducted in a high-risk population.Fair
Mathew et al, 199740

2,069 persons age 35 to 64 years residing in Trivandrum District, India; proportion using tobacco or alcohol NR

Subjects' homes
Presence of homogeneous leukoplakia, ulcerated leukoplakia, verrucous leukoplakia, erythroplakia, nodular leukoplakia, submucous fibrosis, or growths suggestive of oral cancerAdvanced health workers (university graduates) who received 6 weeks of training on oral visual inspection and detection of lesionsPhysician's exam 1 to 6 months after screening examSensitivity=94.3% (200/212)
Specificity=98.3% (1,826/1,857)
PPV=86.6% (200/231)
NPV=99.3% (1,826/1,838)
Selected for re-exam from about 9,000 screened based on population density. Substudy of RCT among general population.37Good
Examination by general dentists
Jullien et al, 199539

United Kingdom
2,027 patients and relatives age ≥40 years of a general dental or medical practice in London; 38% smoked, all drank alcohol

Dental hospital or medical practice
White patch, red patch, or ulcer of >2 weeks duration or presence of specific lesions: lichen planus, lupus erythematosus, submucous fibrosis, or actinic keratosisGeneral dental practitioner, community dental officer, or junior hospital dentistDental specialist at the same visitSensitivity=74% (40/54)
Specificity=99% (1,953/1,973)
PPV=67% (40/60)
NPV=99% (1,953/1,967)
Participation rate NR. Dental clinic patients might be more likely to have oral lesions than general population, and screeners may perform more thorough exams.Good
Downer et al, 199538

United Kingdom
309 persons age ≥40 years employed by a commercial company in London; proportion using tobacco or alcohol NR

Company dental practice
White patch, red patch, or ulcer of >2 weeks durationGeneral dentists with no special trainingOral medicine specialist at same visitSensitivity=71% (12/17)
Specificity=99% (290/292)
PPV=86% (12/14)
NPV=98% (290/295)
53% of those eligible participated; sample enriched from second work site.Fair
Mouth self-examination
Elango et al, 201144

34,766 persons age >10 years residing in Trivandrum District, India (47% ≥40 years); 33% of men and 3% of women used tobacco, betel quid, or alcohol (17.6% total)

Subjects' homes
Presence of white patch, red patch, nonhealing ulcer, difficulty opening mouth, other oral symptomsSubjects performing mouth self-exam as described in brochureHealth worker with 1 month of training on oral cancer conducted exam 4 weeks after screening examSensitivity=18% (39/219)
Specificity=99.9% (34,532/34,547)
PPV=72% (39/54)
NPV=99% (34,532/34,712)
72% of those eligible were examined by health worker.
87% performed mouth self-exam. Only 21% (8/39) with potentially malignant lesions confirmed by health worker followed up at screening clinic
Scott et al, 201045

United Kingdom
53 persons age ≥45 years who smoked (as identified by a general practitioner) residing in London; proportion using alcohol NR

Research department
Presence of ulcer, white patch, red patch, lump, or swellingSubjects performing mouth self-exam as described in a leaflet, after a screening exam by a dentistDentist, at same visitSensitivity=33% (4/12)
Specificity=54% (22/41)
PPV=17% (4/23)
NPV=73% (22//30)
22% of those invited participated. Mouth self-exam conducted in research department after dentist's exam with dentist in the room. Small sample size. Study conducted in a high-risk population.Fair
Toluidine blue
Su et al, 201046

7,975 persons age ≥15 years (61% ≥40 years) who smoked cigarettes or chewed betel quid residing in Keelung, Taiwan; proportion using alcohol NR

Community setting
Presence of any visible lesion including submucous fibrosis, leukoplakia, erythroplakia, lichen planus, ulcer, hyperkeratosis, candidiasisDentists with at least 3 years of practice and additional training; exam with or without toluidine blueIncidence of oral cancer from National Cancer Registry during 4 to 5 years of followupMalignant transformation rate in toluidine blue group: 129 per 100,000 person-years
Malignant transformation rate in placebo dye group: 420 per 100,000 person-years
Relative malignant transformation rate, toluidine blue vs. placebo dye: 0.31 (95% CI, 0.03 to 2.94)
Annual incidence of oral cancer in toluidine blue group: 28.0 per 100,000 person-years
Annual incidence in placebo dye group: 35.4 per 100,000 person-years
Relative incidence of oral cancer, toluidine blue vs. placebo dye: 0.79 (95% CI, 0.24 to 1.23)
78% of those eligible participated. Study conducted in a high-risk population.Fair

Participants who had reference standard.

Abbreviations: NR: not reported; NPV: negative predictive value; PPV: positive predictive value; RCT: randomized, controlled trial

From: 3, RESULTS

Cover of Screening for Oral Cancer
Screening for Oral Cancer: A Targeted Evidence Update for the U.S. Preventive Services Task Force [Internet].
Evidence Syntheses, No. 102.
Olson CM, Burda BU, Beil T, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.