Internet-Delivered Health Interventions That Work: Systematic Review of Meta-Analyses and Evaluation of Website Availability.

BACKGROUND: Due to easy access and low cost, Internet-delivered therapies offer an attractive alternative to improving health. Although numerous websites contain health-related information, finding evidence-based programs (as demonstrated through randomized controlled trials, RCTs) can be challenging. We sought to bridge the divide between the knowledge gained from RCTs and communication of the results by conducting a global systematic review and analyzing the availability of evidence-based Internet health programs. OBJECTIVES: The study aimed to (1) discover the range of health-related topics that are addressed through Internet-delivered interventions, (2) generate a list of current websites used in the trials which demonstrate a health benefit, and (3) identify gaps in the research that may have hindered dissemination. Our focus was on Internet-delivered self-guided health interventions that did not require real-time clinical support. METHODS: A systematic review of meta-analyses was conducted using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (PROSPERO Registration Number CRD42016041258). MEDLINE via Ovid, PsycINFO, Embase, Cochrane Database of Systematic Reviews, and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) were searched. Inclusion criteria included (1) meta-analyses of RCTs, (2) at least one Internet-delivered intervention that measured a health-related outcome, and (3) use of at least one self-guided intervention. We excluded group-based therapies. There were no language restrictions. RESULTS: Of the 363 records identified through the search, 71 meta-analyses met inclusion criteria. Within the 71 meta-analyses, there were 1733 studies that contained 268 unique RCTs which tested self-help interventions. On review of the 268 studies, 21.3% (57/268) had functional websites. These included evidence-based Web programs on substance abuse (alcohol, tobacco, cannabis), mental health (depression, anxiety, post-traumatic stress disorder [PTSD], phobias, panic disorders, obsessive compulsive disorder [OCD]), and on diet and physical activity. There were also evidence-based programs on insomnia, chronic pain, cardiovascular risk, and childhood health problems. These programs tended to be intensive, requiring weeks to months of engagement by the user, often including interaction, personalized and normative feedback, and self-monitoring. English was the most common language, although some were available in Spanish, French, Portuguese, Dutch, German, Norwegian, Finnish, Swedish, and Mandarin. There were several interventions with numbers needed to treat of <5; these included painACTION, Mental Health Online for panic disorders, Deprexis, Triple P Online (TPOL), and U Can POOP Too. Hyperlinks of the sites have been listed. CONCLUSIONS: A wide range of evidence-based Internet programs are currently available for health-related behaviors, as well as disease prevention and treatment. However, the majority of Internet-delivered health interventions found to be efficacious in RCTs do not have websites for general use. Increased efforts to provide mechanisms to host "interventions that work" on the Web and to assist the public in locating these sites are necessary.

Prevalence of human norovirus and Clostridium difficile coinfections in adult hospitalized patients.

OBJECTIVE: Human norovirus (HuNoV) and Clostridium difficile are common causes of infectious gastroenteritis in adults in the US. However, limited information is available regarding HuNoV and C. difficile coinfections. Our study was designed to evaluate the prevalence of HuNoV and C. difficile coinfections among adult patients in a hospital setting and disease symptomatology. STUDY DESIGN AND SETTING: For a cross-sectional analysis, 384 fecal samples were tested for the presence of C. difficile toxins from patients (n=290), whom the provider suspected of C. difficile infections. Subsequent testing was then performed for HuNoV genogroups I and II. Multinomial logistic regression was performed to determine symptoms more frequently associated with coinfections. RESULTS: The final cohort consisted of the following outcome groups: C. difficile (n=196), C. difficile + HuNoV coinfection (n=40), HuNoV only (n=12), and neither (n=136). Coinfected patients were more likely to develop nausea, gas, and abdominal pain and were more likely to seek treatment in the winter season compared with individuals not infected or infected with either pathogen alone. CONCLUSION: Our study revealed that patients with coinfection are more likely to experience certain gastrointestinal symptoms, in particular abdominal pain, suggesting an increased severity of disease symptomatology in coinfected patients.

Gender Differences in Non-Toxigenic Clostridium difficile Colonization and Risk of Subsequent C. difficile Infection.

OBJECTIVES: Previous studies suggest that colonization with non-toxigenic Clostridium difficile may protect against toxigenic C. difficile infection (CDI), yet most of the studies were conducted in men. Therefore, we conducted a study to examine this hypothesis in both genders. METHODS: Patients (n=1492) were classified by disease status at baseline and observed for 1 year. Cox proportional hazards regression was used to evaluate CDI rates within 8 weeks post-baseline (short-term) and from 8 weeks to 1 year (long-term follow-up). RESULTS: During short-term follow-up, CDI rates were 5 times greater in females with non-toxigenic Clostridium difficile compared to females without C. difficile (hazard ratio (HR) = 5.13; 95% CI: 1.47-17.83). The comparable HR in males was 0.44 (95% CI: 0.04-4.43). During long term follow-up, CDI rates were similar in those with non-toxigenic C. difficile and those without C. difficile at baseline, for both females and males. Mortality rates were significantly lower for patients colonized by non-toxigenic C. difficile than those with toxigenic C. difficile at baseline, for both genders combined (HR=0.51; 95% CI: 0.28-0.92) and were similar to those with no C. difficile at baseline (HR=0.78; 95% CI: 0.43-1.41). CONCLUSIONS: There were gender differences in the short-term risk of CDI. Mortality was similar for patients colonized with non-toxigenic C. difficile and patients without C. difficile.

Risk of venous thromboembolism associated with peripherally inserted central catheters: a systematic review and meta-analysis.

BACKGROUND: Peripherally inserted central catheters (PICCs) are associated with an increased risk of venous thromboembolism. However, the size of this risk relative to that associated with other central venous catheters (CVCs) is unknown. We did a systematic review and meta-analysis to compare the risk of venous thromboembolism associated with PICCs versus that associated with other CVCs. METHODS: We searched several databases, including Medline, Embase, Biosis, Cochrane Central Register of Controlled Trials, Conference Papers Index, and Scopus. Additional studies were identified through hand searches of bibliographies and internet searches, and we contacted study authors to obtain unpublished data. All human studies published in full text, abstract, or poster form were eligible for inclusion. All studies were of adult patients aged at least 18 years who underwent insertion of a PICC. Studies were assessed with the Newcastle-Ottawa risk of bias scale. In studies without a comparison group, the pooled frequency of venous thromboembolism was calculated for patients receiving PICCs. In studies comparing PICCs with other CVCs, summary odds ratios (ORs) were calculated with a random effects meta-analysis. FINDINGS: Of the 533 citations identified, 64 studies (12 with a comparison group and 52 without) including 29 503 patients met the eligibility criteria. In the non-comparison studies, the weighted frequency of PICC-related deep vein thrombosis was highest in patients who were critically ill (13·91%, 95% CI 7·68-20·14) and those with cancer (6·67%, 4·69-8·64). Our meta-analysis of 11 studies comparing the risk of deep vein thrombosis related to PICCs with that related to CVCs showed that PICCs were associated with an increased risk of deep vein thrombosis (OR 2·55, 1·54-4·23, p<0·0001) but not pulmonary embolism (no events). With the baseline PICC-related deep vein thrombosis rate of 2·7% and pooled OR of 2·55, the number needed to harm relative to CVCs was 26 (95% CI 13-71). INTERPRETATION: PICCs are associated with a higher risk of deep vein thrombosis than are CVCs, especially in patients who are critically ill or those with a malignancy. The decision to insert PICCs should be guided by weighing of the risk of thrombosis against the benefit provided by these devices. FUNDING: None.

Vascular health, diabetes, APOE and dementia: the Aging, Demographics, and Memory Study.

INTRODUCTION: Evidence from clinical samples and geographically limited population studies suggests that vascular health, diabetes and apolipoprotein epsilon4 (APOE) are associated with dementia. METHODS: A population-based sample of 856 individuals aged 71 years or older from all contiguous regions of the United States received an extensive in-home clinical and neuropsychological assessment in 2001-2003. The relation of hypertension, diabetes, heart disease, stroke, medication usage, and APOE epsilon4 to dementia was modelled using adjusted multivariable logistic regression. RESULTS: Treated stroke (odds ratio [OR] 3.8, 95% confidence interval [CI] 2.0, 7.2), untreated stroke (OR 3.5, 95% CI 1.7, 7.3), and APOE epsilon4 (OR 2.8, 95% CI 1.7, 4.5) all increased the odds of dementia. Treated hypertension was associated with lower odds of dementia (OR 0.5, 95% CI 0.3, 1.0). Diabetes and heart disease were not significantly associated with dementia. A significant interaction was observed between APOE epsilon4 and stroke (P = 0.001). CONCLUSIONS: Data from the first dementia study that is representative of the United States population suggest that stroke, the APOE epsilon4 allele and their interaction are strongly associated with dementia.