Ovarian cancer: is targeted screening possible?
You recently diagnosed a 57-year-old patient with advanced ovarian cancer. In accordance with U.S. Preventive Services Task Force recommendation D (see www.ahrq.gov/clinic/uspstfix.htm), your practice is not to screen patients for ovarian cancer, but you have typically diagnosed it when conducting "case finding" in women with nonspecific complaints. However, you note that the vast majority of the pelvic ultrasounds to rule out ovarian cancer are negative. You wonder if a certain constellation of symptoms would help you to better target patients for ovarian cancer screening in order to diagnose it earlier.
What symptoms are suggestive of ovarian cancer and how sensitive are they?
Using PubMed (www.pubmed.gov) Clinical Queries, you select Diagnosis and Sensitivity and enter "ovarian neoplasm AND symptoms."
The symptom survey instrument that was used in this study has not been validated, and the comparability of the cases and controls is questionable because the controls were 10 years younger and from a different clinical practice. Surveying case patients scheduled for surgical exploration of abdominal masses may lead to overreporting of antecedent symptom presence, severity, and duration. Furthermore, the controls were not linked to ultrasonographic studies to rule out ovarian cancer, although given the low incidence of this tumor (1.5% lifetime risk), this effect is presumably negligible. However, this information would have allowed you to calculate specificity and likelihood ratios for symptom combinations. Symptom combinations are not sensitive enough to be used as a diagnostic clinical screening tool for deciding whom to test further to rule out ovarian cancer. With the presence of three symptoms (bloating, increased abdominal size, and urinary urgency), the sensitivity is 43%, and 57% of the malignancies would have been missed. Although the information relating to symptoms that differentiate benign vs. malignant tumors is interesting, the clinical relevance of this distinction is extremely limited since very few clinicians would defer imaging while proceeding under the assumption that the patient has a benign rather than a malignant ovarian mass.
Patient Preferences and Clinical Decision
You are uncertain how to incorporate this information into your clinical practice. You worry that this research may have created unreasonable expectations among your patients in your ability to diagnose ovarian cancer at earlier stages. You decide to continue to listen to your patients for new, nonspecific, more severe symptoms, and exercise your clinical judgment.
RELATED ARTICLE: B.A. Goff, et al.
Frequency of symptoms of ovarian cancer in women presenting to primary care clinics. (JAMA 291: 2705-12, 2004).
* Design. Case-control study.
* Control selection. Women visiting either of two primary care clinics were asked to complete an anonymous survey about the symptoms experienced over the past year. Subjects were given a list of 20 symptoms typically associated with ovarian cancer (in the categories of pain, eating, abdomen, bladder, bowels, menses, intercourse, and miscellaneous). Severity of symptoms were rated on a 5-point scale and reason for clinic visit was ascertained.
* Case definition. Pathologically verified ovarian cancer. Prior to undergoing surgery to remove an ovarian or pelvic mass, these women filled out an identical survey regarding symptoms.
* Results. In the primary care clinics, 1,709 patients completed the survey, of whom 1,011 had "problem visits" (not for mammography or a general checkup). Of the 128 women with pelvic masses, 44 had malignant epithelial cancers. Compared with women presenting to the clinic for problem visits who did not have irritable bowel syndrome (IBS), women with ovarian cancer had significantly more pelvic pain (41% vs. 26%), abdominal pain (50% vs. 30%), bloating (70% vs. 38%), increased abdominal size (64% vs. 19%), and urinary tract symptoms (55% vs. 32%). Compared with women presenting for problem visits including those with IBS, women with ovarian cancer were significantly more likely to have pelvic pain (odds ratio 2.2); abdominal pain (2.3); difficulty eating (2.5); bloating (3.6); increased abdominal size (7.4); abdominal mass (5.4); urinary urgency (2.5); combination of the three symptoms of bloating, increased abdominal size, and urinary urgency (OR 9.4); and the four symptoms of bloating, increased abdominal size, urinary urgency, and constipation (8.6). Women with ovarian cancer and IBS reported significantly more severe symptoms compared with clinical patients and women with benign masses. Women with ovarian cancer reported symptoms occurring every day compared with clinic patients who typically reported having symptoms two-to-three times per month. Finally, women with benign and malignant masses had a median symptom duration of 6 months, compared with 12-24 months in the IBS and clinic patients.
BY JON O. EBBERT, M.D., AND ERIC G. TANGALOS, M.D.
DR. JON O. EBBERT and DR. ERIC G. TANGALOS are with the Mayo Clinic in Rochester, Minn. To respond to this column or suggest topics for consideration, write to Dr. Ebbert and Dr. Tangalos at our editorial offices or firstname.lastname@example.org.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Mindful Practice|
|Author:||Ebbert, Jon O.; Tangalos, Eric G.|
|Publication:||Internal Medicine News|
|Date:||Aug 15, 2004|
|Previous Article:||Depression not a risk factor for premature ovarian failure.|
|Next Article:||Pharmacologic progress in fibromyalgia care.|