Jacobs et al developed a Risk of Malignancy Index (RMI) for predicting the likely pathology of an ovarian or pelvic mass as benign or malignant.
Parameters:
(1) menopausal status
(2) ultrasound findings suggestive of malignancy
(3) serum CA125 (cancer antigen 125) in U/mL
Criteria for postmenopausal:
(1) amenorrhoea > 1 year, or
(2) women over 50 years of age with previous hysterectomy.
Ultrasound findings suggestive of malignancy:
(1) multi-loculated cysts
(2) evidence of solid areas
(3) evidence of metastases
(4) presence of ascites
(5) bilateral lesions
Parameter |
Finding |
Points |
menopausal status |
pre-menopausal |
1 |
|
post-menopausal |
3 |
ultrasound findings suggestive of malignancy |
none |
0 |
|
1 |
1 |
|
>= 2 |
3 |
risk of malignancy index =
= (points for menopausal status) * (points for ultrasound findings) * (serum CA125 in U/mL)
The derivation of the formula from the equation for the likelihood ratio is given in the Jacobs (1990, Appendix page 929).
Interpretation:
• minimum score: 0
• maximum score: > 10,000
• The performance varies with the cutoff selected.
Performance at difference cutoffs for the RMI
RMI cutoff |
Sensitivity |
Specificity |
Positive PV |
Negative PV |
25 |
100, 97, 92, 97 |
62, 55,61, 60 |
48, 40, 45 |
97, 98 |
50 |
95, 97, 92, 94 |
76, 68, 73, 75 |
56, 49, 56 |
97, 97 |
100 |
85, 87, 90, 77 |
88, 85, 85, 90 |
71, 63, 73 |
97, 92 |
150 |
85, 81, 90, 65 |
94, 85, 88, 91 |
70, 68, 77 |
97, 89 |
200 |
85, 87, 87, 58 |
97, 89, 91, 93 |
75, 73, 78 |
96, 87 |
250 |
78, 76, 87, 54 |
99, 91, 93, 96 |
78, 77, 81 |
96, 86 |
NOTE: data for cutoffs at 75, 80 and 125 not shown.
Limitations:
• Morgante et al (1999) found that the RMI of Tinglustad et al (1996) gave better performance than the RMI of Jacobs et al.
• The prevalence of malignancy in the population would affect performance.
Specialty: Hematology Oncology, Surgery, general, Obstetrics & Gynecology
ICD-10: ,