Skip to main content

Table 2 Main characteristics, results and quality rating of four studies on mammography screening

From: Is single reading with computer-aided detection (CAD) as good as double reading in mammography screening? A systematic review

Author, Year (ref)

Study design, Study period,Population, Readers

Index test (I)

Reference test

Results CI= confidence interval Se= sensitivity Sp=specificity

Study quality, Comments

Gilbert et al., 2008 [71]

Prospective, multicentre 2006-2007

I.1: single reading + CAD, n=28,204

Biopsy of suspected cases or follow-up (not all, though; number not reported)

Cancer detection rate:

Moderate

    

Single reading + CAD: 7.02 /1000.

 
 

Population:

  

Double reading: 7.06/1000.

Restricted generalisability since results were based on single reading +CAD by experienced radiologists.

    

Difference not statistically significant (NS).

 
  

I.2: double reading, n=28,204.

   
 

Initially invited: 68,060 women.

    
    

Recall rate:

Incomplete follow-up, particularly affecting the estimates of sensitivity.

 

Investigated: 28,204.

    
 

Aged 50-70 years (1 % > 70 years).

  

Single reading + CAD: 3.9 %.

 
    

Double reading: 3.4 %.

Scanned analogue mammograms.

    

Difference 0.5 % (95 % CI: 0.3;0.8).

 
 

Readers: radiologists (n=17), specially trained staff (n=10).

    
    

Accuracy:

 
    

Single reading + CAD:

 
    

Se= 87.2 %

 
    

Sp= 96.9 %

 
 

All readers had at least 6 years’ experience and >5000 readings/year

  

Double reading:

 
    

Se= 87.7 %

 
    

Sp= 97.4 %

 
    

Difference in sensitivity:

 
    

0.5 % (95 % CI:

 
    

-7.4;6.6), (NS).

 
    

Difference in specificity 0,5% ( CI not specified but reported NS).

 

Gromet et al., 2008 [69]

Retrospective

I.1: Single reading + CAD

Biopsy and follow-up

Cancer detection rate:

Low

 

Population:

  

Single reading + CAD: 4.2/1000.

Retrospective study (controlled for age and time since last screening).

 

231 221 women

  

Double reading: 4.46/1000 (NS).

 
 

2001-05

n=118,808.

   
  

I.2: Double reading

  

Follow-up time unclear.

 

Readers:

    
     

Screening situation not applicable to European conditions (i.e. recall rate higher than accepted in Europe).

 

Single reading + CAD: specialists in mammography.

    
  

n=112,413.

 

Recall rate:

 
    

Single reading + CAD: 10.6 %.

 
 

Double reading: Specialists in mammography + radiology.

  

Double reading:11.9%.

 
    

Difference statistically significant (p=0.001).

 
     

Invitation procedure and blinded readings unclear.

    

Accuracy:

 
    

Single reading + CAD: Se= 90.4 %

Scanned analogue mammograms.

    

Double reading:

 
    

Se=88.0 %.

 
    

Difference statistically significant.

 
    

Percent of recalled with cancer:

 
    

Single reading + CAD: 3.9%.

 
    

Double reading: 3.7%

(NS).

 

Georgian-Smith et al., 2007 [68]

Prospective

I.1: Single reading + CAD

Biopsy and at least 12 months´ follow-up to detect false negatives.

Cancer detection rate:

Low

 

Study period: 2001-03

  

Single reading +CAD: 2.0/1000.

Screening situation not applicable to European conditions. Invitation procedure not described.

  

n=6381.

 

Double reading: 2.4/1000 (NS).

 
 

Population: 6381 consecutive screening examinations

    
  

I.2: Double reading

   
    

Recall rate:

Population, selection criteria, withdrawals unclear.

  

n=6381.

 

Single reading +CAD: 7.87%.

 
    

Double reading: 7.93% (NS).

 
 

Readers:

   

Not independent double reading but blinded to CAD

 

Experienced breast radiologists

  

Accuracy:

 
    

Sensitivity and specificity not reported.

Number of recalls based on all readings.

 

Single reading + CAD.

   

Scanned analogue radiographs.

 

Double reading: Not independent reading.

    

Khoo et al., 2005 [70]

Prospective

I.1: Single reading +CAD n= 6111.

Biopsy

Cancer detection rate:

Low

 

Study period: not reported.

 

Not reported

Total for double reading + single reading + symptomatic patients:10/1000.

A so-called relative sensitivity used since 3-year follow-up not yet achieved.

   

No follow-up

  
 

Population: 6,111 women (45-94 years), screening every 3rd year

    
    

Not reported individually for the groups.

 
     

Relatively high screening age and long screening intervals.

  

I.2: Double reading n= 6111.

   
    

Recall rate:

 
    

Single reading + CAD: 6.1%.

Unclear whether the readings were blinded.

    

Double reading: 5.0 %.

Incomplete follow-up.

 

Readers:

  

Difference statistically significant

Scanned analogue radiographs.

 

Radiologists (n=7) and specially trained staff (n=5).

    
    

Accuracy: (relative sensitivity)*

 
    

Single reading + CAD: Se= 91.5%.

 
    

Double reading: Se= 98.4% (NS).

 
 

Double reading not always performed by two radiologists.

    
  1. * Relative sensitivity= number of detected cancer cases per reader divided by all detected cancer cases (due to lack of follow-up).