Table 3.

Descriptive summary of quality, performance indicators, and case detection rates

Author, year, country,Study design MMAT quality scoreDescription of intervention (study objective, target population, screening test, MSU type, time period of study)Sample (number of clients screened or exams, demographics)Reported cancer screening performance indicatorsCancer detection rate/1,000 exams
Breast cancerAl Mulhim et al. (2015), Saudi Arabia (69)Quantitative descriptive MMAT: 50%Target: 53,800 women ≥ age 40 sampled from 15 primary care centers in Eastern Saudi Arabia Test: digital MMG; MSU type: 2 vans Period: 2009–2014 (5 years); opportunistic screen8,061 clients screened; mean age (SD): 46.5 (9.4); age range: 37–78Abnormal call rate: 7.9% (n = 636); biopsy rate: 0.7% (n = 63) Detected cases: n = 47; test PPV=7.3%a; biopsy PPV=74.6%a Proportion DCIS: 15%, 70.2% of cases <2 cm5.9 (first round); 4.1 (later rounds)
Apffelstaedt et al. (2014), South Africa (31)Quantitative descriptive MMAT: 50%Target: women ≥ age 40 served by primary health centers in Capetown Test: digital MMG; MSU type: 1 van Period: 2011–2012 (2 years); opportunistic screen2,712 clients screenedAbnormal call rate: 9.6% (n = 261); biopsy rate: 1.2% (n = 32) Lost to follow-up among recalls: 4% (n = 11) Detected cases: n = 10 (5 DCIS, 2 stage II; 3 stage III) Test PPV = 3.8%a; biopsy PPV =31%a3.7
Hughes et al. (2014), Australia (13)Retrospective cohort MMAT: 75%Objective: to compare diagnostic effectiveness of MSU-MMG + step-down MSU assessment versus fixed-center (FC)-MMG + diagnostic follow-up Target: women ages 40–69 in Western Australia Test: MSU-MMG + “step down” follow-up of diagnostic views in MSU with referral to urban FC if biopsy is required; MSU type: 1 van Comparison: FC-MMG + FC diagnostic follow-up Period: 1999–2008 (10 years); organized programMSU, FC group: number of exams 214,328, 545,699Results reported for MSU, FC: screening uptake rate: 59%, 57%; repeat participation rates (within 27 mos.): 64.9% (64.7–65.1), 68.3% (68.2–68.4) Tumor size: 17.1 mm, 17.4 mm Interval cancer rates per 10,000 women: ≤12 months: 0.16 (0.02–0.34), 0.70 (0.46–0.94) 13–24 months: 0.54 (0.18–0.89), 0.76 (0.49–1.03) Sensitivity = 95%, 91%3.1 (2.8–3.3) MSU group vs. 7.04 (6.82–7.27) FC group
Renck et al. (2014), Brazil (16)Cross-sectional analytic MMAT: 75%Target: Women ≥ age 40 in southern Brazil Test: MSU-MMG; MSU type: 1 van Comparison: FC-MMG Period: unspecified; opportunistic screenMSU, FC group: clients screened; 8,607, 1,312; mean age (SD): 51.2 (10.4), 51.4 (10.4); first-screen: 48%, 26%Results reported for MSU, FC: Detected cases: n = 37, n = 9, odds of breast cancer detection is similar between MSU and FC: crude OR: 1.60 (CI: 0.77–3.32); adjusted OR for sociodemographic factors (1.56, CI: 0.73–3.31).4.3 (MSU) 6.9 (FC)
Carkaci et al. (2013), USA (17)Quantitative descriptive MMAT: 75%Target: women ≥35 years in Texas Test: film MMG (2009–10), digital MMG (2011–12); MSU type: 1 van Period: 2009–2012 (4 years); opportunistic screen12,726 exams; 33% medically underserved populationAbnormal call rate = 16.2% (film) 15.8% (digital) Detected cases: not reported Operational measures: mean number of exams per day = 23 (film), 20 (digital); Mean daily kilometers = 42 km Total cost per MU = $427,103 (2013 USD)2.3 (film, 2009), 2.9 (digital, 2011)
Fontenoy et al. (2013), Canada (14)Retrospective cohort MMAT: 100%Objective: to compare diagnostic performance of MSUs to FCs Target: Women ages 50–69 in Québec Test: MSU-MMG (digital and film); MSU type: 2 vans, 1 portable unit transport by boat/plane Comparison: FC-MMG Period: 2002–2010 (9 years); results reported for 2002–2006; organized programMSU, FC group: 16,654 exams, 1,000,706 exams; mean age (SD) = 57.9 (5.5), 58.0 (5.5)MSUs coverage rate: 44.3% (2002–05), 63.4% (2006–10); first-screen (MSU): 11% Results of MSU, FC: abnormal call rate: 8.1% (n = 1,354), 9.5% (n = 95,440) Detected cases: n = 72, n = 5,166 Test sensitivity: 69.9%, 70.3%; Test PPV: 5.3%, 5.4% *No differences in validity measures between MSU and FC (compared using adjusted risk ratios with 95% CI)4.3 (MSU) 5.2 (FC)
Brooks et al. (2013), USA (42)Quantitative descriptive MMAT: 75%Target: women ages 40–75 in Louisville, KY at “high risk” (regions with high cancer mortality and poverty) Test: digital MMG and CBE; MSU type: 1 van Period: 2008–2010 (3 years); opportunistic screen3,923 clients screened (4,542 MMG, 4,516 CBE); mean age = 54.6; 29% ≥ 5 years since last examAbnormal call rate: 9.3% (n = 424)a Detected cases: n = 31; Test PPV: 7.3% (31/424)a Factors associated with abnormal exam: age (<50), no insurance, Hispanic ethnicity, smoking, and family history7.9
Haikel et al. (2012), Brazil (30)Quantitative descriptive MMAT: 75%Target: women ages 40–69 in São Paulo state Test: film MMG; MSU type: multiphasic van (also offered Pap test for cervical cancer) Period: 2003–2005 (3 years); opportunistic screen10,521 clients screened; mean age (SD) = 51 (7.8); First-screen: 42%Abnormal call rate: 9.4%; biopsy rate: 2.6% Detected cases: n = 41; Test PPV: 4.5%; Biopsy PPV: 16% Operational measures: mean daily number of exams = 26.33.9
Rodriguez-Cuevas et al. (2009), Mexico (71)Quantitative descriptive MMAT: 75%Target: Women ages ≥40 residing in Mexico City's Federal District Test: film MMG; MSU type: 7 vans Period: 2005–2006 (2 years); opportunistic screen96,828 clients screened; Mean age (SD) = 49.3 (8.5)Abnormal call rate: 1% (n = 949) Loss to follow-up among abnormal exams: 21% Detected cases: n = 208; Test PPV = 27.7% (208/949)a Proportion of cases detected in situ or stage I = 29.4%2.1
Del Turco et al. (2007), Italy (18)Retrospective cohort MMAT: 75%Study objective: compare performance measures of digital and film MMGs (both offered on MSU) Target: Women ages 50–69 residing in Florence Test group: digital MMG; Comparison group: film MMG (matched by age and radiologist) MSU type: 2 vans Period: 2004–2005 (2 years); organized programDigital, film group: Clients screened: n = 14,385, n = 14 385 First-screen: 12.2%, 11.8%Results of digital, film group: Abnormal call rate: 4.3% (n = 618), 3.5% (n = 498) Recall rate (poor quality): 0.3% (n = 39), 0.5% (n = 72) Detected cases: n = 104, n = 84 Test PPV: 15.9%, 14.7% Proportion DCIS: 27.8% (n = 29), 17.8% (n = 15) χ2 tests showed difference between groups*Digital, film: 7.4, 8.2 (first round) 7.2, 5.5 (repeat rounds)
Zappa et al. (2002), Italy (76)Quantitative descriptive MMAT: 75%Study objective: To monitor the occurrence of interval breast cancers Target: Women ages 40–70 (1970–1990); ages 50–70 (post 1990) in the district of Florence Test: film MMG; MSU type: 1 van Period: 1985–1993 (9 years); organized program80,134 exams: (13,153 first round, 66,981 repeat rounds)Detected cases: n = 277 Proportion of screen-detected in-situ: 6.5% (18/277) 90 incidence cases detected ≤3 years after negative exam Breast density was associated with increased risk of developing interval case4.9 (first round) 3.17 (repeat rounds)
Gordenne et al. (2000), Belgium (78)Quantitative descriptive MMAT: 50%Target: 99,699 women ages 40–69 residing in the province of Liège Test: film MMG; MSU type: 3 vans Period: 1992–1998 (8 years); organized program52,475 exams (29,918 clients screened)Mean attendance rate = 25%; abnormal call rate: 3.3% (n = 1,012); recall rate (poor quality): 0.7% (n = 343) Detected cases: n = 207; Test PPV: 20.5%a Proportion DCIS: 16% (n = 30); tumors <10 mm: 36% (n = 70)3.9a
Frelix et al. (1999), USA (35)Quantitative descriptive MMAT: 75%Target: women ≥40 in The Bronx borough of New York city (underserved population) Test: film MMG + CBE; MSU type: 2 vans Period: 1994–95 (16 mos.); opportunistic screen1,944 clients screened; 45.6% uninsuredAbnormal call rate: 18.5% (n = 360); biopsy rate: 2.5% (n = 49) Compliance rate among biopsy recommendations: 100% Detected cases: n = 26; test PPV: 6.9%a; biopsy PPV: 51% Proportion DCIS: 16% (n = 4); stage I: 36% (n = 9)13.3a
Kann et al. (1998), USA (82)Quantitative descriptive MMAT: 75%Target: Women ≥40 in Long Island, New York Test: film MMG and CBE; MSU type: 1 van Period: 1990–1994 (5 years); organized program19,280 examsBiopsy recommendation rate: 1.5% (n = 295); biopsies performed: 0.8% (n = 157); Loss to follow-up: 46.8% (n = 138) Detected cases: n = 64; PPV recommended & performed biopsies: 21.6% & 40.8%a Operational measures: 30 tests offered per day3.3a
Horton et al. (1996), UK (66)Prospective cohort MMAT: 50%Target: women ages 65–74 in Hamstead, UK, not routinely invited to breast cancer screening through the National Health Service (NHS) Test: film MMG; MSU type: 1 van Comparison: women ages 50–64 invited for screening by the NHS Period: 1992–1993 (0.5 year); organized program4,576 clients screenedResults for age groups: 50–64, 65–69, 70–74 respectively Abnormal call rate: 3.2% (n = 92), 4.1% (n = 32), 4.8% (n = 44) Biopsy rate: 0.5% (n = 14), 1.6% (n = 12), 1.5% (n = 14) Detected cases: n = 13, n = 11, n = 12 Test PPV: 14.1%, 34.4%, 27.3%a Biopsy PPV: 92.9%, 91.7%, 85.7%a4.5 (50–64 years; 14.2 (65–69 years); 13.2 (70–74 years)
Séguret et al. (1995), France (72)Quantitative descriptive MMAT: 100%Target: 52,617 women ages 40–70 in Herault region, France Test: film MMG; MSU type: 1 van Period: 1990–1992 (2.5 years); organized program26,026 clients screenedScreening coverage rate = 48% Abnormal call rate: 7% (n = 1,826); loss to follow up among recalls = 10.6%; Biopsy rate: 1.5% (n = 393) Detected cases: n = 137; reported PPV (8.4%) based on clients who returned for follow-up; Test PPV: 7.5%a; Biopsy PPV: 35%; Proportion cases in situ: 18% (n = 27); tumors <1 cm: 42%; Detected interval cancers within 30 mos.: n = 135.3
Van Oyen et al. (1994), Belgium (74)Quantitative descriptive MMAT: 75%Target: 57,339 women ages 50–64 in Antwerp and Limburg provinces Test: MMG not otherwise specified MSU type: 1 semi-mobile unit (mammography machine, transported in an adapted trailer and set up in a central clinic in each visited town) Period: 1989–1992 (4 years); organized program16,017 clients screenedScreening coverage rate = 28%; Abnormal call rate: 4.1% (n = 652); Loss to follow up among abnormal calls: 15% Biopsy rate: 0.6% (n = 89) Detected cases: n = 46; Test PPV: 8.3% (reported); Test PPV: 0.071a; Biopsy PPV: 51.7% Proportion DCIS: 30%; stage 1 (<2 cm): 59.5% Median time from screening to diagnosis: 4 weeks Operational measures: 45 exams offered daily2.9
Garas et al. (1994), Greece (73)Quantitative descriptive MMAT: 100%Target: 42, 411 women ages 40–64 in the Ilia and Messinia counties, southern Greece Test: film MMG; MSU type: 1 van (multimodal, also provided Pap tests, results unreported) Period: 4 years; organized program22,258 clients screenedScreening coverage rate = 52.5%; Abnormal call rate: 5.3% (n = 1169); biopsy rate: 0.8% (n = 176) (recommended), 0.7% (n = 158) performed, loss to follow-up: 11% Detected cases: n = 69; test PPV: 7.4%; biopsy PPV: 47.1% Proportion DCIS: 4.3%; stage 1 (< 2cm): 51.5%3.1
Dershaw et al. (1992), USA (55)Quantitative descriptive MMAT: 75%Target: women age ≥35 in New York state Test: film MMG (Pap tests also offered, results not reported by study) MSU type: 1 van, 2 semi-mobile MMG units Period: 1988–1991 (3 years); opportunistic screen22,540 examsAbnormal call rate: 11.2% (n = 2515); biopsy recommended rate: 0.6% (n = 288); biopsy performed rate Detected cases: n = 50; test PPV: 2.0%; biopsy PPV (recommended): 17.3%; biopsy PPV (performed): 21% Operational measures: 25 exams offered daily. Costs (USD 1992): average cost per test = $65; startup costs $360,000;2.2
McCoy et al. (1992), USA (36)Quantitative descriptive MMAT: 50%Target: women age ≥40 in Florida (socioeconomically disadvantaged) Test: film MMG (Pap tests also offered, results not reported by study); MSU type: 1 van Period: 1987–1991 (4 years); opportunistic screen12,456 clients screened; First-screen: 72%Abnormal call rate: 21.3% (n = 2,660) Detected cases: n = 90, test PPV: 4.0%a Proportion of cases DCIS: 10% (years 1987–89), 24% (years 1990–91)6.6 (prevalent rate) 5.5 (overall rate)a
Tabar and Gad (1981), Sweden (22)RCT MMAT: 100%Target: 47,000 women age ≥40 Ostergotland county, Sweden; MSU type: 1 van Test: film MMG; control: no screening (22,000 women in Kopparberg county) Period: 1977–1980 (3 years); organized program34,187 clients screenedScreening coverage rate among intervention group: 84.3% Abnormal call rate: 4.8% (n = 1655); biopsy rate: 0.1% (n = 362) Detected cases: n = 235; Test PPV: 14.2%a; biopsy PPV: 64.9%a6.9a
Strax (1973), USA (23)RCT MMAT: 50%Target: 31,000 women age 40–64 participating in the Health Insurance Plan (HIP) in New York City Test: film MMG and CBE; MSU type: 1 van control group: no screening (31,000 women randomly matched on age, through HIP program) Period: 1963–1969 (7 years); organized program20,211 clients screenedIntervention group: screening coverage rate: 66% Detected cases: n = 296; detection rate per 1,000 PYs: 2.72 70% of cases were without axillary nodal involvement Comparison group: Detected cases: n = 284; detection rate per 1,000 PYs: 1.86 45% of cases were without axillary nodal involvement2.3
Cervical CancerLorenzi et al. (2016), Brazil (21)Quantitative descriptive MMAT: 75%Objective: To evaluate the efficiency of the careHPV test in detecting high-risk HPV in women with no precursor lesions and evaluate performance as a primary screening method Target: women ages 18–85 residing in 4 states of South-East and Central-West Brazil Cotesting: CareHPV + Pap-liquid based cytology (Pap-LBC) MSU type: 1 van Period: 03–12/2012 (8 mos.); opportunistic screen3,068 clients screened; Median age (range) = 47 (18–85)Test positive for HR-HPV: 10.0% (n = 307) Abnormal cytology: total 4.3% (n = 132): [ASC/AGC-US: n = 66, ASC/AGC-H: n = 13, LSIL: n = 38, HSIL: n = 15]; HR-HPV positivity among cytology samples: 8.2% of normal samples, 39.4% of ASC/AGC-US, 38.5% of ASC/AGC-H, 55.3% of LSIL, 100% of HSIL Colposcopy rate: 10.8% referred (n = 332); 3.6% performed (n = 111); 66% loss to follow-up Biopsy rate: 2.1% (n = 66), CIN1: 40.9% (n = 27); CIN2: 7.6% (n = 5); CIN3: 6.1% (n = 4); invasive carcinoma: 4.5% (n = 3) HPV test: Sensitivity: 100% (CI: 75.3–100%), specificity: 10.8% (CI: 5.1–19.6%)1 invasive cancer/1,000 women cotested
Paengchit et al. (2014), Thailand (25)Quantitative descriptive study MMAT: 100%Objective: describe prevalence and genotype distribution of HPV in Lampang, Thailand Target: women ages 30–70 Test: Pap-LBC, HR-HPV testing; MSU type: 1 van Period: 01–03/2013 (3 mos.); opportunistic screen2,000 clients screened; Mean age (range) = 47.8 (30–70)Test positive for HR-HPV: 5.4% (CI: 4.5–6.5; n = 108) Rate of abnormal cytology: 1.95% (1.4–2.7) Number with HSIL or higher: n = 19 HR-HPV positivity among cytology samples: 4.0%, 60.0%, 89.5% among normal, ASC-US/LSIL, HSIL respectively9.5 HSIL/1,000 smearsa
Fregnani et al. (2013), Brazil (27)Cross-sectional analytic MMAT: 100%Test: Pap, comparison: Pap LBC MSU type: 1 van Period: 05–12/2010 (8 mos.); opportunistic screenPap, Pap-LBC Clients screened: 6,047, 6,001 Mean age (SD): 46.1 (13.1), 46.5 (12.5)Results for Pap, Pap-LBC: Abnormal call rate: 0.021, 0.01; Specimen adequacy rate: 99.92%, 99.97% ASCUS-US: 0.1% (n = 6), 0.7% (n = 39); ASC-H: 0.3% (n = 21), 0.4% (n = 24); LSIL: 0.3% (n = 19) vs 0.7% (n = 41); HSIL: 0.2% (n = 14) vs 0.4% (n = 22)Rate of HSIL per 1,000 smears: 2 vs. 4
Swaddiwud-hipong et al (1999), Thailand (20)Mixed: qualitative case-study + quantitative descriptive MMAT: 100%Objective: measure effect of MSU program on knowledge and attitudes toward screening Target: low-income women residing in rural regions of Thailand, ages 18–65 Test: Pap; MSU type: 1 van; Period: 1992–1996 (5 years); Opportunistic screen13,081 clients screened; Women surveyed in 1991, 1994, 1997: n = 1,603, n = 1,369, n = 1,576LSIL: n = 46 (0.4%); HSIL (%): n = 46 (0.4%); detected cases: n = 7 Survey results from 1991, 1994, 1997 Self-reported knowledge of Pap test: 21%, 57%, 76% Self-reported use of Pap test: 20%, 58%, 70%0.5 cases/1,000 smears 3.5 HSIL/1,000 smears
Megevand et al. (1996), South Africa (15)Prospective cohort MMAT: 100%Objective: to identify operational strategies to minimize loss to follow-up Target: low-income women in Capetown Test: (phase I) Pap + Colposcopy referral to FC; (phase II “See, screen, treat”): Pap + Colposcopy in MSU; MSU type: 1 van; Period: 1993 (8 mos.); opportunistic screenPhase I, II: clients screened; 2,619, 2,426 Mean age (range): 34 (19–83), 31 (17–78); Total First-screen: 75.2%Results for Phase 1, 2: Abnormal Pap smears - LSIL: 3.3% (n = 87), 3.5% (n = 87); HSIL: 3.2% (n = 86), 1.3% (n = 33) Effective completion of colposcopy among HSIL: 33% (29/86), 97% (32/33) Operational measures: 50 tests offered per day0.92/1,000 smearsa
Thornton et al. (1989), UK (54)Quantitative descriptive MMAT: 100%Target: Women >40 employed by worksites in West Surrey/North-East Hampshire region (UK) Test: Pap, CBE (unreported); MSU type: 1 van Period: 1985–1986 (9 mos.); opportunistic screen568 clients screenedCoverage rate among women in targeted workplaces: 91% Abnormal call rate: n = 64 (0.11); Detected cases: 1; Test PPV: 1.5%a1.76 cases/1,000 smears
Brindle et al. (1976), UK (49)Quantitative descriptive MMAT: 100%Target: women (age unspecified) residing in the UK; Test: Pap; MSU type: 1 van Period: 1973 (3 mos.); opportunistic screen1,526 clients screened; First-screen: 70%Abnormal call rate: n = 10 (0.66%)6.6 abnormal/1,000 smears
Whitfield et al (1972), UK (50)Quantitative descriptive MMAT: N/ATarget: women (16–60) residing in the UK Test: Pap (CBE also offered, results unreported); MSU type: 1 van Period: 1969–71 (3 years); opportunistic screen1,952 clients screened; first-screen: 70%Number of cases carcinoma in situ: n = 4 Cost measurements (reported in 1972 pounds (£)): Cost per test = £2, average cost per detected CIN3 = £2002.04 cases of carcinoma in situ/1,000 smears
Breast, cervicalMauad et al. (2009), Brazil (70)Quantitative descriptive MMAT: 100%Target: 54,238 women ages 40–69 for breast screening, 117,239 women ages 20–69 for cervical screening Test: Film MMG and Pap MSU type: 1 van Period: 2003–2004 (2 years); opportunistic screenMMG, Pap: clients screened: 7,192, 2,964; First-screen: 44%, 7%Breast: screening coverage rate: 13%; Abnormal call rate: 7.6% (n = 549); biopsy rate: 1.4% (n = 105); Detected cases: n = 22; test PPV: 4%a; biopsy PPV: 20.9%a Cervical: screening coverage rate: 2.5% Abnormal call rate: 1.9% (n = 59)a; cytologic abnormalities: 0.5% (n = 15): ASCUS 0.1% (n = 3); CIN1 0.13% (n = 4); CIN2 0.1% (n = 3); CIN3 0.1% (n = 3); invasive squamous cell carcinoma 0.07% (n = 2); operational measures: 40 exams offered dailyBreast: 3.1 cases/1,000 exams Cervical: 0.7 cases/1,000 exams
Breast, cervical, prostate, gastric, oral, colon and lungLynch (1976), USA (94)Quantitative descriptive MMAT: 50%Objective: Measure effectiveness of MSU by comparing survival with general U.S. population Target: Adults residing in rural Nebraska; Test: unspecified; MSU type: custom-built house, 6 m × 18 m in dimension (transported using tractor) Period: 1971–1975 (5 years); opportunistic screen5,232 clients screened (1,984 men and 3,248 women); Mean age (range) = 55 (18–90)Detected cases: N = 22 (8 breast, 7 colon, 2 endometrium, 1 lung, 1 prostate, 1 penis, 1 lip, 1 stomach) 1 cancer diagnosis per 238 clients screened Survival: observed vs. expected deaths: 3 vs. 7.5 (estimated using U.S. cancer mortality rates)4.2 cases/1,000 people screened
Breast, cervical, prostate, gastric, oral, colon skin, penileLynch, (1973), USA (67)Quantitative descriptive MMAT: 25%Target: individuals residing in eight communities in Nebraska; Test: MMG (NOS), CBE, Pap, DVI, proctosigmoidoscopy, OVE, laryngoscopy MSU type: custom-built house (transported by tractor) Period: 1971–1972 (2 years); opportunistic screen3,040 screening exams; Mean age = 60 (males), 56 (females)Total number of malignancies: n = 13 (n = 64 including skin) Operational measures: 16 patients scheduled per hour4.3 cases/1,000 people screened
  • Abbreviations: CBE, clinical breast exam; CI, 95% confidence interval; DRE, digital rectal exam; DVI, direct visual inspection; FC, fixed-centre; Pap, Papanicolau smear with conventional cytology; Pap-LBC, Papanicolau smear with liquid-based cytology; NOS, not otherwise specified; OVE, oral visual examination; VILI, visual inspection with Lugol iodine.

  • aDenotes hand calculations. The objective is left unspecified for studies which are purely descriptive. If a distinct objective outside of program description was mentioned, this objective was stated. See Supplementary Table S3 for reference to the MMAT scale.