Physician's Prescribing Preference as an Instrumental Variable: Exploring Assumptions Using Survey Data.

BACKGROUND: Physician's prescribing preference is increasingly used as an instrumental variable in studies of therapeutic effects. However, differences in prescribing patterns among physicians may reflect differences in preferences or in case-mix. Furthermore, there is debate regarding the possible assumptions for point estimation using physician's preference as an instrument. METHODS: A survey was sent to general practitioners (GPs) in The Netherlands, the United Kingdom, New Zealand, Ireland, Switzerland, and Germany, asking whether they would prescribe levothyroxine to eight fictitious patients with subclinical hypothyroidism. We investigated (1) whether variation in physician's preference was observable and to what extent it was explained by characteristics of GPs and their patient populations and (2) whether the data were compatible with deterministic and stochastic monotonicity assumptions. RESULTS: Levothyroxine prescriptions varied substantially among the 526 responding GPs. Between-GP variance in levothyroxine prescriptions (logit scale) was 9.9 (95% confidence interval: 8.0, 12) in the initial mixed effects logistic model, 8.3 (6.7, 10) after adding a fixed effect for country and 8.2 (6.6, 10) after adding GP characteristics. The occurring prescription patterns falsified the deterministic monotonicity assumption. All cases in all countries were more likely to receive levothyroxine if a different case of the same GP received levothyroxine, which is compatible with the stochastic monotonicity assumption. The data were incompatible with this assumption for a different definition of the instrument. CONCLUSIONS: Our study supports the existence of physician's preference as a determinant in treatment decisions. Deterministic monotonicity will generally not be plausible for physician's preference as an instrument. Depending on the definition of the instrument, stochastic monotonicity may be plausible.

Instrumental variable analysis as a complementary analysis in studies of adverse effects: venous thromboembolism and second-generation versus third-generation oral contraceptives.

PURPOSE: A potentially useful role for instrumental variable (IV) analysis may be as a complementary analysis to assess the presence of confounding when studying adverse drug effects. There has been discussion on whether the observed increased risk of venous thromboembolism (VTE) for third-generation oral contraceptives versus second-generation oral contraceptives could be (partially) attributed to confounding. We investigated how prescribing preference IV estimates compare with conventional estimates. METHODS: Women in the Clinical Practice Research Database who started a second-generation or third-generation oral contraceptive from 1989 to 2013 were included. Ordinary least squares and two-stage least squares regression were used to estimate risk differences in VTE. Cox regression and IV for Cox proportional hazards regression were used to calculate hazard ratios (HR). The instrument used was the proportion of prescriptions for third-generation oral contraceptives by the general practitioner in the year preceding the current prescription. RESULTS: All analyses pointed in the direction of an increased VTE risk for third-generation oral contraceptives. The adjusted HR from the conventional Cox regression was 1.62 (95% confidence interval 1.16-2.27) and the fully adjusted HR from the IV Cox regression was 3.45 (95% confidence interval; 0.97-11.7), showing a larger risk and wider confidence intervals in the IV analysis. CONCLUSIONS: The similarity in direction of results from the IV analyses and conventional analyses suggests that major confounding is unlikely. IV analysis can be a useful complementary analysis to assess the presence of confounding in studies of adverse drug effects in very large databases.

Physician's preference-based instrumental variable analysis: is it valid and useful in a moderate-sized study?

BACKGROUND: Instrumental variable methods can potentially circumvent the unmeasured confounding inherent in observational data analyses. METHODS: We investigated the validity and usefulness of physician's preference instrumental variable analysis in the setting of a moderate-sized clinical study. Using routine care data from 476 elective cardiac surgery patients, we assessed the effect of preoperative corticosteroids on mechanical ventilation time and duration of intensive care and hospital stay, occurrence of infections, atrial fibrillation, heart failure, and delirium. RESULTS: Although results of the physician's preference-based instrumental variable analysis corresponded in direction to results of a recent large randomized trial of the same therapy, the instrumental variable estimates showed much larger effects with very wide confidence intervals. CONCLUSION: The lesser statistical precision limits the usefulness of instrumental variable analysis in a study that might be of sufficient size for conventional analyses, even if a strong and plausible instrument is available.

Sex-Related Differences in the Immune Response to Meningococcal Vaccinations During Adolescence.

Background: Immune responses to pediatric vaccinations have been reported to differ according to sex. Such sex-differential responses may become more pronounced during adolescence due to hormonal differences. We investigated whether the vaccine response following primary vaccination against meningococcal serogroup A (MenA), MenW and MenY and booster vaccination against MenC differed between girls and boys using data from two clinical studies. Methods: Children aged 10, 12, and 15 years, who had been primed with MenC vaccination between 14 months and 6 years of age, received a booster MenC vaccination or MenACWY vaccination. Polysaccharide-specific IgG concentrations and functional antibody titers [determined with the serum bactericidal antibody (SBA) assay] were measured at baseline, 1 month, 1 year, and 3 years (only MenC group) after vaccination. We calculated geometric mean concentrations and titers (GMC and GMT) ratios for girls vs. boys adjusted for age group. Additionally, we compared the proportion protected individuals between girls and boys at all timepoints. Results: This study included 342 girls and 327 boys from two clinical trials. While MenAWY antibody levels did not differ consistently 1 month after vaccination, all GMC- and GMT-ratios were in favor of girls 1 year after vaccination [range: 1.31 (1.02-1.70) for MenA IgG to 1.54 (1.10-2.16) for MenW IgG]. Overall, MenC antibody levels were slightly higher in girls at all postvaccination timepoints (GMC- and GMT-ratios: 1.16/1.17 at 1 month, 1.16/1.22 at 1 year and 1.12/1.15 3 years postvaccination). Higher MenC antibody levels were observed in 12- and 15-year-old girls compared to boys of the same age, whereas 10-year-old boys and girls had similar antibody levels. The percentage of participants protected (SBA titer ≥ 8) was very high (95-100%) at all timepoints, and did not differ significantly between boys and girls. Conclusion: Antibody responses were higher in girls than in boys for all serogroups at most timepoints after primary MenAWY vaccination and booster MenC vaccination. The differences in average titers were however small and the percentage participants with protective titers was very high for both sexes.

Viral load dynamics in intubated patients with COVID-19 admitted to the intensive care unit.

BACKGROUND: Prolonged viral RNA detection in respiratory samples from patients with COVID-19 has been described, but the clinical relevance remains unclear. We studied the dynamics of SARS-CoV-2 on a group and individual level in intubated ICU patients. METHODS: In a cohort of 86 patients, we analysed SARS-CoV-2 RT-PCR results on nasopharyngeal and sputum samples (obtained as part of clinical care twice a week) according to time after intubation. Subsequently, we performed survival analyses. RESULTS: 870 samples were tested by RT-PCR. Overall viral load was highest in the first week (median nasopharynx 3.5, IQR 1.5-4.3; median sputum 4.3, IQR 3.3-5.6) and decreased over time. In 20% of patients a relapsing pattern was observed. Nasopharyngeal and sputum PCR status on day 14 was not significantly associated with survival up to day 60 in this small cohort. CONCLUSION: In general SARS-CoV-2 RNA levels in respiratory samples in patients with severe COVID-19 decrease after the first week after intubation, but individual SARS-CoV-2 RNA levels can show a relapsing pattern. Larger studies are needed to address the association of clearance of SARS-CoV-2 RNA from respiratory samples with survival, because we observed a trend towards better survival in patients with early clearance from sputum.

Socioeconomic Status Is Associated With Antibody Levels Against Vaccine Preventable Diseases in the Netherlands.

Background: We investigated whether low socioeconomic status (SES), which is associated with reduced health and life expectancy, might play a role in increased risk for infectious diseases. Therefore, we explored the association between SES and immunoglobulin G (IgG) levels against various pathogens. Methods: We analyzed the association between SES [educational level and net household income (NHI)] and serum IgG concentration against measles, mumps, rubella, varicella, Haemophilus influenzae type B (HiB), pneumococcus, meningococcus serogroup C (MenC), and cytomegalovirus (CMV) collected within a national cross-sectional serosurvey (2006/2007) using linear regression analyses among non-vaccinated individuals. Results: Higher educational level was associated with higher IgG concentrations against measles (GMC ratio 1.34, 95% CI 1.18-1.53) and rubella (1.13, 1.02-1.25) compared to low education level. In contrast, higher education level was associated with lower IgG concentrations against pneumococcus (0.78, 0.70-0.88), MenC (0.54, 0.44-0.68), and CMV (0.23, 0.18-0.31) compared to low education level. This pattern was also evident when NHI was used as SES indicator. Conclusion: Our study suggests that socioeconomic status is associated with antibody levels in a pathogen-dependent manner. The results suggest that differences in serological response upon infection or differences in exposure might be involved in the variation in IgG levels between SES groups.

Differences by sex in IgG levels following infant and childhood vaccinations: An individual participant data meta-analysis of vaccination studies.

BACKGROUND: If immune responses to vaccination differ between males and females, sex-specific vaccination schedules may be indicated. We systematically reanalysed childhood vaccination studies conducted in The Netherlands for sex-differences in IgG-responses. To assess the impact of potential sex-differences in IgG-responses, we explored sex-differences in vaccine failure/effectiveness and reactogenicity. METHODS: Six studies with IgG-measurements for 1577 children following infant pneumococcal (PCV7/PCV10/PCV13) and/or DTaP-IPV-Hib(-HepB) vaccinations, or the pre-school DTaP-IPV booster were included. We performed one-stage individual participant data meta-analyses per time-point of the effect of sex on IgG levels against pneumococcal serotypes, diphtheria toxoid, tetanus toxoid, pertussis Ptx/FHA/Prn and Hib-PRP using linear mixed models. Using existing study data, we compared reactogenicity after PCV7/PCV10 and DTaP-IPV-Hib(-HepB) vaccination in girls and boys. Vaccine failure/effectiveness was compared between girls and boys for invasive pneumococcal disease (IPD), invasive Hib disease and pertussis using notification data. RESULTS: For pneumococcal vaccination, the geometric mean concentration ratio of IgG levels in girls versus boys pooled across serotypes was 1.15 (95%CI 0.91-1.45) 1 month following the primary series, 1.16 (1.02-1.32) at age 8 months, 1.12 (1.02-1.23) pre-booster (age 11 months) and 0.99 (0.89-1.10) post-booster (age 12 months). Diphtheria toxoid, tetanus toxoid, pertussis Ptx/FHA/Prn and Hib-PRP IgG levels did not differ between girls and boys, except for Hib post-booster (1.24; 95%CI 1.01-1.52) and tetanus before pre-school booster (0.71; 0.53-0.95). We found no difference between boys and girls in reactogenicity at age 4 or 11 months or in vaccine failure/effectiveness for IPD, invasive Hib disease or pertussis. CONCLUSION: For most vaccine antigens investigated, there were no consistent differences in vaccine-induced IgG levels. Vaccine-induced pneumococcal IgG levels were slightly higher in girls, but only between the primary series and the 11-month booster. These results, along with similar reactogenicity and vaccine failure/effectiveness, support the uniform infant vaccination schedule in the Dutch national immunisation programme.

Non-specific effects of measles, mumps, and rubella (MMR) vaccination in high income setting: population based cohort study in the Netherlands.

Objectives To investigate whether measles, mumps, and rubella (MMR) vaccine has positive non-specific effects in a high income setting and to compare rates of hospital admissions for infections between children aged ≤2 years who received live MMR vaccine and those who received an inactivated vaccine against diphtheria, tetanus, pertussis, polio, and Haemophilus influenzae type b (DTaP-IPV-Hib) as their most recent vaccination.Design Nationwide population based cohort study.Setting In the Netherlands, DTaP-IPV-Hib+pneumococcal vaccination (PCV) is recommended at ages 2, 3, 4, and 11 months and MMR + meningococcal C (MenC) vaccination at age 14 months. Data from the national vaccine register were linked to hospital admission data.Participants 1 096 594 children born in 2005-11 who received the first four DTaP-IPV-Hib+PCV vaccines.Main outcome measures Hazard ratio for admission to hospital for infection in children with MMR+MenC compared with the fourth DTaP-IPV-Hib+PCV as their most recent vaccination. Cox regression was performed with most recent vaccination as time dependent variable, adjusted for potential confounders. Analyses were repeated with admission for injuries or poisoning as a negative control outcome. In addition, rate of admission for infection was compared between the fourth and third DTaP-IPV-Hib+PCVas most recent vaccination.Results Having had MMR+MenC as the most recent vaccination was associated with a hazard ratio of 0.62 (95% confidence interval 0.57 to 0.67) for admission to hospital for infection and 0.84 (0.73 to 0.96) for injuries or poisoning, compared with the fourth DTaP-IPV-Hib+PCV as most recent vaccination. The fourth DTaP-IPV-Hib+PCV as most recent vaccination was associated with a hazard ratio of 0.69 (0.63 to 0.76) for admission to hospital for infection, compared with the third DTaP-IPV-Hib+PCV as most recent vaccination.Conclusions Healthy vaccinee bias could at least partly explain the observed lower rate of admission to hospital with infection after MMR vaccination. The lower rate is associated with receipt of any additional vaccine, not specifically MMR vaccine. This emphasises the caution required in the interpretation of findings from observational studies on non-specific effects of vaccination.

Mendelian randomization studies: a review of the approaches used and the quality of reporting.

BACKGROUND: Mendelian randomization (MR) studies investigate the effect of genetic variation in levels of an exposure on an outcome, thereby using genetic variation as an instrumental variable (IV). We provide a meta-epidemiological overview of the methodological approaches used in MR studies, and evaluate the discussion of MR assumptions and reporting of statistical methods. METHODS: We searched PubMed, Medline, Embase and Web of Science for MR studies up to December 2013. We assessed (i) the MR approach used; (ii) whether the plausibility of MR assumptions was discussed; and (iii) whether the statistical methods used were reported adequately. RESULTS: Of 99 studies using data from one study population, 32 used genetic information as a proxy for the exposure without further estimation, 44 performed a formal IV analysis, 7 compared the observed with the expected genotype-outcome association, and 1 used both the latter two approaches. The 80 studies using data from multiple study populations used many different approaches to combine the data; 52 of these studies used some form of IV analysis; 44% of studies discussed the plausibility of all three MR assumptions in their study. Statistical methods used for IV analysis were insufficiently described in 14% of studies. CONCLUSIONS: Most MR studies either use the genotype as a proxy for exposure without further estimation or perform an IV analysis. The discussion of underlying assumptions and reporting of statistical methods for IV analysis are frequently insufficient. Studies using data from multiple study populations are further complicated by the combination of data or estimates. We provide a checklist for the reporting of MR studies.

Sample size importantly limits the usefulness of instrumental variable methods, depending on instrument strength and level of confounding.

OBJECTIVES: Instrumental variable (IV) analysis is promising for estimation of therapeutic effects from observational data as it can circumvent unmeasured confounding. However, even if IV assumptions hold, IV analyses will not necessarily provide an estimate closer to the true effect than conventional analyses as this depends on the estimates' bias and variance. We investigated how estimates from standard regression (ordinary least squares [OLS]) and IV (two-stage least squares) regression compare on mean squared error (MSE). STUDY DESIGN: We derived an equation for approximation of the threshold sample size, above which IV estimates have a smaller MSE than OLS estimates. Next, we performed simulations, varying sample size, instrument strength, and level of unmeasured confounding. IV assumptions were fulfilled by design. RESULTS: Although biased, OLS estimates were closer on average to the true effect than IV estimates at small sample sizes because of their smaller variance. The threshold sample size above which IV analysis outperforms OLS regression depends on instrument strength and strength of unmeasured confounding but will usually be large given the typical moderate instrument strength in medical research. CONCLUSION: IV methods are of most value in large studies if considerable unmeasured confounding is likely and a strong and plausible instrument is available.

[Instrumental variable analysis]. / Instrumentele-variabele-analyse.

Instrumental variable analysis is a recently propagated method to deal with confounding and to estimate therapeutic effects in observational studies. An instrumental variable is a factor which affects treatment but is not related to patient prognosis. A theoretical advantage of a well-chosen instrumental variable is that both measured and unmeasured confounders do not influence the effect estimator. Examples of instrumental variables previously used are regional differences in treatment and physician prescribing preference. Application of instrumental variable analysis seems most suited to large patient registries with considerable expected residual confounding in case standard analytical methods are applied.

Low C-reactive protein levels in a traditional West-African population living in a malaria endemic area.

BACKGROUND: C-reactive protein (CRP) levels are reported to be elevated in populations of African descent living in affluent environments compared to populations of European ancestry. However, the natural history of CRP levels in populations of African descent living under adverse environments remains largely unknown. METHODS: CRP levels were measured with a high sensitivity assay in 624 apparently healthy individuals who contributed blood as part of a study on innate immune responsiveness in a traditional Ghanaian population living under adverse environmental conditions in a malaria endemic area. As a comparison, we included CRP measurements from 2931 apparently healthy individuals from the Dutch population that were included in the same batch of CRP analyses. Associations between CRP and body mass index (BMI), immune responsiveness, and P. falciparum parasitaemia were investigated. RESULTS: In an age- and sex-adjusted model, CRP levels were 0.54 mg/L lower in the Ghanaian compared to the Dutch cohort (1.52 vs. 0.98 mg/L, p<0.001). When accounting for the substantially higher average BMI in the Dutch compared to the Ghanaians (25.6 vs. 18.4 kg/m(2)) the difference in CRP levels disappeared. BMI associated positively with CRP in the Dutch but not in the Ghanaians. In individuals with an acute phase response, CRP levels were higher in the Ghanaian compared to the Dutch cohort (24.6 vs. 17.3 mg/L, p = 0.04). Levels of CRP were positively related to immune responsiveness and P. falciparum parasitaemia (all p<0.001) among Ghanaians. CONCLUSIONS: Our study demonstrates that West-Africans do not exhibit an inherently high inflammatory state. The role of genes, environment and gene-environment interaction in explaining reports of elevated CRP levels in populations of African ancestry when compared to other ethnicities living in affluent environments thus merits further investigation.

Parasitic infections and immune function: effect of helminth infections in a malaria endemic area.

According to the hygiene hypothesis, reduced exposure to infections could explain the rise of atopic diseases in high-income countries. Helminths are hypothesised to alter the host's immune response in order to avoid elimination and, as a consequence, also reduce the host responsiveness to potential allergens. To elucidate the effect of current helminth infections on immune responsiveness in humans, we measured cytokine production in a rural Ghanaian population in an area with multiple endemic parasites including malaria, intestinal helminths and protozoa. Multiplex real-time PCR in stool samples was used for the detection of four gastrointestinal helminths, of which only Necator americanus was commonly present. A similar assay was used to test for Giardia lamblia in stool samples and malaria infection in venous blood samples. Levels of the cytokines interleukin (IL)-10, tumour necrosis factor (TNF)-α, IL-17, IL-6, IL-13, and interferon (IFN)-γ were determined in whole-blood samples ex vivo-stimulated either with lipopolysaccharide (LPS) and zymosan (for innate cytokine production) or the T-cell mitogen phytohaemagglutinin (PHA). There were no significant differences in either innate or PHA-stimulated cytokine production dependent on current N. americanus infection. Plasmodium falciparum malarial infection was associated with a pro-inflammatory response indicated by increased innate production of TNF-α, IL-17 and IL-6. There was no clear pattern in cytokine responses dependent on G. lamblia-infection. In conclusion, in this rural Ghanaian population current N. americanus infections are not associated with altered immune function, while infection with P. falciparum is associated with pro-inflammatory innate immune responses.