2 Minute Medicine Rewind October 21, 2019

Gender differences in Twitter use and influence among health policy and health services researchers

1. Women’s voices on common social media platforms appear less influential and have less reach than their male counterparts in the health policy and health service research arena, suggesting that social media may re-enforce known gender disparities in academic medicine.

Evidence Rating Level: 3 (Average)

It has been shown that women in academic medicine experience lower visibility and academic success than their male counterparts. The objective of this cross-sectional analysis of publicly available data was to describe gender differences in Twitter use and influence among health service researchers. Using a sampling group of 6,442 speakers and co-authors of research presented at AcademyHealth’s 2018 Annual Research Meeting, 3,148 health service researchers were identified (1,668 females and 1,480 males) using online searches, of which 919 (492 females and 427 males) used Twitter. Twitter’s application program interface was used to extract metrics regarding Twitter use and reach for the last 3,200 tweets. A natural language processing model was used to predict the gender of users’ followers and those they followed. Outcomes of interest were number and gender of users’ followers and those they followed, and influence of users’ tweets in terms of likes and re-tweets. Study findings showed that women were more likely to have female followers (54.8% of users followed by women were women, versus 42.6% followed by men). Women were also found to have substantially less influence on Twitter as well, with half the mean number of followers (567.5 (SD 1819.7) vs. 1162.3 (3056.2), p<0.001), and their tweets generating fewer annual mean likes (315.6 (SD 659.8) vs. 577.6 (1281.8)) and re-tweets (207.4 (SD 403.6) vs. 399.8 (876.6)). Likes and re-tweets per tweet were also significantly lower amongst women compared to men (likes: 3.8 (SD 4.8) vs. 4.5 (4.8); re-tweets: 2.4 (SD 2.2) vs. 3.1 (3.4)). Gender differences held across all levels of Twitter activity, and were generally largest among full professors. The findings of this study therefore suggest that while social media is often considered an equitable platform to represent both men and women in academic medicine, women’s voices on Twitter appear less influential and have less reach than their male counterparts, potentially re-enforcing gender disparities.

Patterns of work-related burnout in physician-scientists receiving career development awards from the National Institutes of Health

1. Increased competing demands and a more negatively-perceived work climate may contribute to increased rates of burnout amongst female physician-scientists relative to their male counterparts.

Evidence Rating Level: 3 (Average)

Physician burnout is increasingly prevalent, and is more common among certain subgroups, including women. The objective of this national survey was to examine patterns of work-related burnout specifically amongst physician-scientists. Recipients of the K08 and K23 awards from the National Institutes of Health between 2006 and 2009 were surveyed between August 2010 and February 2011, with a follow-up survey administered to respondents in 2014. Responses from physician-respondents who remained in academic positions at the time of follow-up were analysed to evaluate baseline individual, job and environmental variables that were associated with work-related burnout, measured using the Copenhagen Burnout Inventory subscale. Of 816 participants, men were more likely to consider their work climate as being positive than women (49.4% vs. 35.8%, p<0.001). Women reported higher weekly hours committed to concurrent parenting and domestic tasks (>=44 hours by 41.7% of women vs. 17.1% of men, p<0.001), and lower weekly hours of patient care (>=15 hours by 14.2% vs. 24.1%, p=0.007) and work overall (>=60 hours by 34.6% vs. 59.2%, p<0.001). Mean Copenhagen Burnout Inventory scores were 45.3 (SD 14.2) and 42.2 (14.4) between women and men, respectively, with burnout identified in 41.4% and 31.5% in both groups (p=0.004). Greater time spent on parenting and domestic tasks, less vacation time, increased time spent on patient care and a negatively perceived work climate were associated with burnout. Contrary to bivariate analysis, sex was not associated with greater burnout on multivariate analysis. The findings of this study therefore suggest that increased competing demands and more negative perceptions of work climate drive much of the differences in burnout between female and male physician-scientists, and that interventions to address these areas may contribute to significantly alleviating burnout.

Association of early postnatal transfer and birth outside a tertiary hospital with mortality and severe brain injury in extremely preterm infants: observational cohort study with propensity score matching

1. Extremely preterm infants born in non-tertiary hospitals have an increased risk of death, and transfer in the first 48 hours postnatally is associated with increased risk of severe brain injury relative to infants born in tertiary neonatal care.

Evidence Rating Level: 2 (Good)

Extremely preterm infants born at less than 28 weeks of gestation are at an increased risk of death and neonatal morbidity. When women at risk of extremely preterm delivery present to non-tertiary hospitals, one may consider transfer to a tertiary care setting before delivery, immediately after delivery (postnatally) or delayed after delivery following complete stabilization of the infant. The introduction of a networked model of neonatal care in England has allowed for increased early postnatal transfer of these infants to tertiary care settings, which may, in fact, be associated with adverse outcomes. This retrospective cohort study aimed to examine how early postnatal transfer compares to ongoing non-tertiary neonatal care with respect to mortality and severe brain injury as outcomes among extremely preterm infants. This study used data on all extremely preterm infants admitted to neonatal units in England between January 2008 and December 2015, collected via the UK National Neonatal Research Database. Infants were grouped into four categories: the upward transfer group, with those born in hospitals with local neonatal units and transferred within 48 hours; the non-tertiary group, with those born in such hospitals and not transferred within 48 hours; the horizontal transfer group, with those born in a tertiary hospital and transferred to a different tertiary hospital for non-clinical reasons within 48 hours due to non-clinical reasons (e.g. insufficient capacity); and the control group, with infants born in a tertiary hospital and not transferred within 48 hours. Propensity score matching and pre-defined background variables were used to create subgroups with equal distributions of confounding variables. Of 2,181 infants included (727 from each of the upward transfer, non-tertiary and control groups matched), infants in the upward transfer group were found to have similar odds of death before discharge (OR 1.22, 95% CI 0.92 to 1.61) but significantly higher odds of severe brain injury (OR 2.32, 95% CI 1.78 to 3.06, NNT 8) and lower odds of survival without severe brain injury (OR 0.60, 95% CI 0.47 to 0.76, NNT 9) compared to controls. The non-tertiary care group was found to have significantly higher odds of death (OR 1.34, 95% CI 1.02 to 1.77, NNT 20) but similar odds of severe brain injury (OR 0.95, 95% CI 0.70 to 1.30) and survival without severe brain injury (OR 0.82, 95% CI 0.64 to 1.05) compared to controls. There was no significant difference in death before discharge between upward transfer and non-tertiary groups, but a significantly lower risk of severe brain injury (OR 0.41, 95% CI 0.31 to 0.53, NNT 8) and higher odds of survival without severe brain injury (OR 1.37, 95% CI 1.09 to 1.73, NNT 14) were reported. The findings of this study therefore suggests that extremely preterm infants born in non-tertiary hospitals may be at an increased risk of death, and that transfer in the first 48 hours postnatally may be associated with an increased risk of severe brain injury relative to infants born in tertiary neonatal care without transfers.

Efficacy of antibiotic treatment in patients with chronic low back pain and Modic changes (the AIM study): double blind, randomised, placebo controlled, multicentre trial

1. A three-month course of amoxicillin provides no clinically significant benefit over placebo for treatment of chronic back pain with previous disc herniation and type 1 or 2 Modic changes.

Evidence Rating Level: 1 (Excellent)

As current management strategies for chronic low back pain offer low to moderate improvement in pain and disability, efforts to identify subgroups of patients who may benefit from specific treatments are underway. There is some evidence to suggest that chronic back pain with endplate changes in vertebral bone marrow (Modic changes) may reflect low-grade bacterial discitis by Cutibacterium acnes via bloodstream source, and that antibiotic treatment may be effective in reducing pain and disability in this subgroup. The AIM study examined the efficacy and harm of three months of oral amoxicillin therapy for patients with chronic low back pain and type 1 or 2 Modic changes. Using a multi-centre, double-blind, parallel group design with a three-month treatment phase and a nine-month follow-up phase, adults with chronic low back pain of at least moderate intensity, lumbar disc herniation on recent magnetic resonance imaging, and type 1 or type 2 Modic changes were recruited from outpatient clinics from six Norwegian hospitals between 2015 and 2017. Participants (n=180) were randomized to three months of oral amoxicillin 750 mg three times daily or placebo (maize starch). The primary outcome of interest was the validated Norwegian Roland-Morris Disability Questionnaire (RMDQ) score at one-year follow-up in an intention-to-treat population. The minimal clinically important between-group difference was pre-defined at 4 for this measure. Researchers found a non-clinically important difference in the amoxicillin treatment group at one-year follow-up in terms of mean RMDQ scores (mean difference -1.6, 95% CI -3.1 to 0.0, p=0.04). Adverse events occurred in 56% of patients in the treatment group, compared to 34% in the placebo group. The findings of this study therefore suggest that three-month amoxicillin therapy in chronic low back pain patients with previous disc herniation and Modic changes do not provide a clinically-important benefit compared to placebo.

Weight change across adulthood in relation to all cause and cause specific mortality: prospective cohort study

1. Maintaining obesity across adulthood, gaining weight from young to middle adulthood, and losing weight from middle to late adulthood, are all associated with an increased risk of all-cause mortality among community-dwelling adults.

Evidence Rating Level: 2 (Good)

While it is known that high body mass index (BMI) is associated with higher risk of premature death, evidence regarding weight change from young to late adulthood, or middle to late adulthood, in terms of mortality is inconsistent and equivocal. Using data from the US National

Health and Nutrition Examination Survey (NHANES), this study aimed to examine the association between weight change from young adulthood to midlife and to late adulthood  (defined as ages 25 years, 47 years and 57 years, respectively) with all-cause and cause-specific mortality. NHANES was a prospective cohort study, which collected nationally representative health data between 1988 and 2014. This analysis included participants age 40 years and over at baseline, with 40,052 participants identified and 36,051 participants finally included in the analysis. Measured body weight and height at baseline and recalled weights at young adulthood and middle adulthood stages before baseline were gathered. Primary outcomes of interest were all-cause and cause-specific mortality from baseline to the end of December 2015. Of 10,500 deaths over a 12.3 year mean follow-up period, those moving from non-obese to obese categories between young and middle adulthood had a 22% higher risk of all-cause mortality (HR 1.22, 95% CI 1.11 to 1.33), and 49% had a higher risk of cardiac-specific mortality (HR 1.49, 95% CI 1.21 to 1.83), compared to those who remained at a normal weight. Changes from obese to non-obese BMI was not associated with significant differences in mortality risk in the same life periods, but were associated with an increased risk of all-cause mortality (HR 1.30, 95% CI 1.16 to 1.45) and cardiac-specific mortality (HR 1.48, 95% CI 1.14 to 1.92) from middle to late adulthood; moving from non-obese to obese categories between these latter life periods was not significantly associated with increased mortality risk. Maintaining obesity throughout adulthood was consistently associated with increased risk of all-cause mortality for all life periods. The findings of this study therefore suggest that obesity throughout adulthood, weight gain from young to middle adulthood and weight loss from middle to late adulthood are associated with an increased risk of mortality; maintaining normal weight throughout adulthood is likely associated with the greatest benefit.

Image: PD