EBM Focus - Year in Review 2014

EBM Focus - Year in Review 2014

Read the full year in review 2014 and earn CME credit.Pelvic Exam Not Recommended for Screening Asymptomatic Nonpregnant Women

Reference: Ann Intern Med 2014;161:67

Pelvic examination is a common part of annual health visits for women as a screening method for cancer and infections. A new evidence-based clinical practice guideline by the American College of Physicians (ACP) found that the harms outweigh the potential benefits, and recommends strongly against screening pelvic examinations in asymptomatic nonpregnant adult women.

The guideline included a systematic review of English-language studies published from 1946 to 2014 to determine the best current evidence for the diagnostic accuracy, benefits, and harms of the screening pelvic examination for detection of cancer, pelvic inflammatory disease, or other benign gynecologic conditions. For cancer detection, across 3 cohort studies with 5,633 asymptomatic women, only 4 cases of ovarian cancer were detected by pelvic examination. And in a randomized trial evaluating screening for ovarian cancer with transvaginal ultrasonography and CA-125 in 78,000 women, the bimanual pelvic examination was dropped after 5 years because no cancer was detected solely by this examination. The ACP is clear in distinguishing the pelvic exam from cervical cancer screening using visual inspection and swabs for cervical cytology, and cervical cancer screening can be completed without a bimanual pelvic examination.

No studies were identified that evaluated the diagnostic accuracy or benefits of pelvic examination for asymptomatic pelvic inflammatory disease, benign conditions, or gynecologic cancer other than cervical or ovarian cancer. Harms identified for the pelvic examination included pain or discomfort reported by 11%-60% in 8 studies with 4,576 women, and fear, embarrassment, or anxiety reported in 10%-80% in 7 studies with 10,702 women. Women experiencing pain or discomfort during their examination were less likely to have a return visit than those not experiencing pain or discomfort in 5 of 5 studies reporting this outcome.

Many aspects of routine care are based on established practice, and have not been subject to rigorous assessment or review. For cancer in particular, there has been a focus in mainstream health messaging that early detection is vital, and many people may not understand that screening carries potential harms that must be weighed against the proposed benefits and may not understand that some of our “standard approaches” can have little or no efficacy for providing early detection that leads to clinical benefit. This is the first evidence-based guideline evaluating screening pelvic examination in this patient population, and the recommendation against this practice will have an important impact on annual wellness visits for a very large number of women.

For more information see the Ovarian cancer screening topic in DynaMed.

Arthroscopic Partial Meniscectomy Does Not Improve Symptoms of Degenerative Medial Meniscus Tear in Patients Without Knee Osteoarthritis

Reference:N Engl J Med 2013 Dec 26;369(26):2515 (level 1 [likely reliable] evidence)

Arthroscopic surgeries for patients with established knee osteoarthritis are becoming less common due to a lack of clinical evidence supporting their use. A Cochrane review found that arthroscopic surgery is ineffective for unselected patients with knee osteoarthritis (Cochrane Database Syst Rev 2008 Jan 23;(1):CD005118), based partly on the findings of a randomized trial showing no significant improvement in pain or function scores with either arthroscopic debridement or arthroscopic lavage compared to placebo surgery (N Engl J Med 2002 Jul 11;347(2):81 full-text). Another randomized trial subsequently showed that the addition of arthroscopic surgery to physical and medical therapies did not improve function or pain scores in patients with moderate-to-severe knee osteoarthritis (N Engl J Med 2008 Sep 11;359(11):1097 full-text). More recently, a randomized trial in patients with meniscal tear and knee osteoarthritis showed that arthroscopic partial meniscectomy plus physical therapy did not improve symptoms more than physical therapy alone (N Engl J Med 2013 May 2;368(18):1675 full-text, see DynaMed EBM Focus Volume 8, Issue 13).

However, the implications for patients without clearly established knee osteoarthritis have remained unclear. A recent randomized trial comparing meniscectomy to strengthening exercises in patients presenting with degenerative medial meniscus tear and no clear evidence of osteoarthritis (Kellgren-Lawrence grade 0-1) found no significant between-group differences in function, pain, or patient satisfaction scores (Am J Sports Med 2013 Jul;41(7):1565). Now, a randomized trial compares arthroscopic partial meniscectomy to sham surgery in patients with medial meniscus tear without knee osteoarthritis.

A total of 146 patients aged 35-65 years with symptomatic degenerative medial meniscus tear without knee osteoarthritis were randomized to arthroscopic partial meniscectomy vs. sham surgery and followed for 12 months. Postoperative care, including walking aids and instructions for graduated exercises, was identical for both groups, and all patients were instructed to take over-the-counter analgesics as required. Symptoms were assessed using the Lysholm and Western Ontario Meniscal Evaluation Tool (WOMET) scores, which both range from 0 to 100, with higher scores indicating less severe symptoms. Knee pain was assessed after exercise using a numeric rating scale with a range of 0 to 10, with higher scores indicating greater pain severity.

Both groups had a significant improvement from baseline in symptom and knee pain scores, but there were no significant between-group differences for these outcomes at 12 months. The mean improvement in the Lysholm score was 21.7 points with arthroscopic partial meniscectomy group vs. 23.3 points with sham surgery, with a difference of 11.5 points considered clinically meaningful. Similarly, the mean improvement in the WOMET score was 24.6 points with partial meniscectomy vs. 27.1 points with sham surgery, with a difference of 15.5 points considered clinically meaningful. The mean improvement in knee pain was 3.1 points with partial meniscectomy vs. 3.3 points with sham surgery, with a difference of 2 points considered clinically meaningful. There were also no significant differences in the rate of subsequent knee surgery, patient-reported satisfaction, patient-reported improvement, or serious adverse events.

Recent clinical evidence from randomized trials has consistently shown a lack of efficacy of arthroscopic surgeries for patients with knee osteoarthritis, including patients with meniscus tears. This trial extends those findings to patients with meniscus tears but without established knee osteoarthritis, showing no significant difference in symptom or pain scores between patients receiving arthroscopic partial meniscectomy and those receiving sham surgery. It should be noted that the Lysholm and WOMET scores used in this trial for symptom assessment are primarily focused on basic activities of daily living, and may not adequately describe high-level function as may be required for strenuous work or sports.

For more information, see the Meniscus tears topic in DynaMed.

Maternal Vaccination Decreases Risk of Influenza in Mothers and Their Infants

Reference: N Engl J Med 2014 Sep 4;371(10):918
for women without HIV infection (level 1 [likely reliable] evidence)
for women with HIV infection (level 2 [mid-level] evidence)

Pregnant women are at increased risk of severe influenza from their second trimester until the early postpartum period, and are considered a priority group for receiving seasonal influenza vaccine by the World Health Organization (Weekly Epidemiological Record PDF) and the Centers for Disease Control and Prevention (MMWR Recomm Rep 2013 Sep 20;62(RR-07):1 full-text). However, data on the efficacy of influenza vaccination in this population is limited, particularly data showing a protective effect on infants after birth. Two recent randomized trials from South Africa evaluated the efficacy of trivalent inactivated influenza vaccination in 2,116 pregnant women not infected with HIV, in 194 pregnant women with HIV infection, and in their newborns up to 24 weeks after birth.

In pregnant women without HIV infection, the rate of laboratory-confirmed influenza was 1.8% with influenza vaccine vs. 3.6% with placebo (p = 0.01). Similarly, among infants born to vaccinated mothers, confirmed influenza infection rates were 1.9% with influenza vaccine vs. 3.6% with placebo (p = 0.01). In pregnant women infected with HIV, the rates of laboratory-confirmed influenza with vaccine vs. placebo were 7% vs. 17% (p = 0.05) in mothers and 5% vs. 6.8% among their infants (not significant). There were no significant differences between the influenza group and the placebo group for rates of nonspecific influenza-like illness or for respiratory illnesses in general, irrespective of HIV infection status.

The protective effect of inactivated influenza vaccine for infants of vaccinated mothers observed in these new trials is consistent with a previous trial conducted in Bangladesh. This previous trial of 340 women found that a similar maternal vaccination reduced laboratory-confirmed influenza cases in infants followed for 24 weeks compared to maternal vaccination with pneumococcal vaccine (N Engl J Med 2008 Oct 9;359(15):1555). This newer and larger trial further solidifies this protective effect in mothers without HIV infection and their infants. It also suggests benefit for HIV-infected mothers as well, a population that may be particularly vulnerable to infection.

For more information see the Influenza in pregnancy topic in DynaMed.

Bilateral Mastectomy May Not Increase Survival Compared to Breast-Conserving Surgery With Radiation in Women With Unilateral Breast Cancer

Reference: JAMA 2014 Sep 3;312(9):902 (level 2 [mid-level] evidence)

Bilateral mastectomy rates have been on the rise among women with early breast cancer in recent years (J Clin Oncol 2009 Sep 1;27(25):4-82, J Clin Oncol 2011 Jun 1;29(16):2158). However, while bilateral mastectomy has been shown to reduce the incidence of breast cancer in women at increased risk (Ann Oncol 2013 Aug;24(8):2029 full-text), the evidence is inconsistent for bilateral mastectomy as a treatment in women with unilateral breast cancer (Cochrane Database Syst Rev 2010 Nov 10;(11):CD002748). Furthermore, no randomized trials have been performed comparing bilateral mastectomy vs. breast-conserving surgery plus radiation. Many women with breast cancer have a preference for bilateral mastectomy based on its perceived benefits, and may object to randomization to less extensive surgery, making it difficult to conduct randomized trials for this particular comparison. A new population-based cohort study evaluated bilateral mastectomy, unilateral mastectomy, and breast-conserving surgery with radiation in 189,734 women with early (stage 0-III) unilateral breast cancer using data collected from 1998 to 2011 in the California Cancer Registry.

A total of 6.2% of women had bilateral mastectomy, 38.8% had unilateral mastectomy, and 55% had breast-conserving surgery with radiation. The overall rate of bilateral mastectomy increased during the study, ranging from 2% in 1998 to 12.3% in 2011. Factors most significantly associated with increased rate of bilateral mastectomy vs. breast-conserving surgery with radiation included age

< 50 years, positive lymph node status, and lobular histology. However, data on other relevant information that may have influenced a treatment decision (such as magnetic resonance imaging or BRCA1 or BRCA2 gene mutation status) were not available from the registry.

There were no significant differences in the all-cause or breast cancer-specific mortality comparing bilateral mastectomy vs. breast-conserving surgery with radiation. Both approaches were associated with small but significant reductions in mortality compared to unilateral mastectomy. Estimated 10-year all-cause mortality was 16.8% with breast-conserving surgery with radiation, 18.8% with bilateral mastectomy, and 20.1% with unilateral mastectomy.

This study supports the use of breast-conserving surgery with radiation as the first line of treatment for women with early stage unilateral breast cancer. However, the lack of information on potential confounders, such as magnetic resonance imaging or gene mutation results, detailed patient history, and physician recommendations limit the ability to make definitive conclusions from the current data.

For more information see the

Surgery for early and locally advanced breast cancer topic in DynaMed.

In Adults With Unruptured Brain Arteriovenous Malformation, Interventional Therapy Appears to Worsen Outcomes Compared to Medical Management

Reference: Lancet 2014 Feb 15;383(9917):614 (level 2 [mid-level] evidence)

With the increased use of noninvasive neuroimaging, there has been an increase in the detection of brain arteriovenous malformations prior to symptomatic bleeding, but there is currently no clear consensus for the management of these lesions. Patients diagnosed with unruptured or asymptomatic arteriovenous malformations may be managed conservatively or, alternatively, they may be offered interventional therapy with the aim of obliterating the origin of the arteriovenous malformation. Several interventional therapies, including neurosurgery, embolization, and stereotactic radiotherapy have been used successfully in these patients, but there is little clinical evidence to guide the choice of interventional therapy, or to demonstrate its superiority over conservative management. A prospective population-based cohort study in Scotland previously showed that patients receiving interventional therapy for arteriovenous malformation had worse outcomes than those who did not (Lancet Neurol 2008 Mar;7(3):223). Now, a randomized trial compares the addition of interventional therapy to medical management vs. medical management alone in adults with unruptured arteriovenous malformation.

A total of 226 adults (mean age 45 years) with unruptured brain arteriovenous malformation were randomized to medical management alone vs. medical management plus interventional therapy. Patients, clinicians, and investigators were aware of treatment assignment. Medical management consisted of pharmacologic therapy for existing medical disorders (such as seizures or headaches) or any coexisting vascular risk factors (such as diabetes or arterial hypertension), as needed. Patients randomized to interventional therapy additionally received neurosurgery, embolization, or stereotactic radiotherapy, either alone or in combination, at the discretion of the local trial investigator. Patients with evidence of recent or prior hemorrhage, or who had received previous interventional therapy for brain arteriovenous malformation, were excluded from the trial. The goal of the interventional therapy was complete eradication of the arteriovenous malformation. The primary outcome was a composite of symptomatic stroke or death.

The trial was stopped early by an independent data and safety monitoring board, due to superiority of medical management alone based on a prespecified stopping value for the primary outcome. The interim analysis included 223 patients (99% of those randomized) with a median follow-up of 33 months. Symptomatic stroke or death occurred in 10% of those receiving medical management alone vs. 31% of those receiving interventional therapy plus medical management (p < 0.05, NNT 5). The incidence of stroke was 11% with medical management alone vs. 39% with interventional therapy plus medical management (p < 0.0001, NNT 4). In addition, neurologic deficits unrelated to stroke occurred in 0.9% of those treated with medical management alone vs. 12% of those treated with interventional therapy plus medical management (p = 0.0008, NNT 9).

The results of this trial extend previous findings from an observational study, and further support the conclusion that current interventional therapies for unruptured arteriovenous malformations do not improve cerebrovascular outcomes. However, arteriovenous malformations are associated with a long natural history, and the median follow-up of 33 months included in this trial limits the ability to make conclusions about long-term effectiveness. In addition, by not including patients who had received previous interventional therapy for brain arteriovenous malformation, patients with a more aggressive disease course may have been excluded, and this may therefore limit the generalizability of the findings.

For more information see the Central nervous system vascular malformations topic in DynaMed.