Random Intercept Linear Models Research Articles

Household food insecurity and future orientation of Ghanaian youth and their parents

ABSTRACTFood insecurity heightens risk for poor mental health and psychosocial functioning. Higher levels of future orientation influence numerous desirable behaviors. However, limited evidence exists on the association between food insecurity and future orientation, particularly in low-resource settings. The objective of this study was to examine the association between food insecurity and future orientation of Ghanaian youth and their parents. The study included a cross-sectional sample of 2656 youth and 2656 parents from 8 of Ghana’s 10 regions. Food insecurity was measured using the Household Food Insecurity Access Scale. Future orientation in the Ghanaian context was measured using three distinct factors adapted from the School Success Profile and the Consideration of Future Consequences scale. We analyzed our data using hierarchical linear modeling, with a three-level linear random-intercept model with covariates. Results suggest an inverse relationship between food insecurity and future orientation of youth and their parents. Food insecurity was consistently and significantly associated with lower orientation toward success (β = −0.18, 95% confidence interval [CI] = −0.22, −0.14) and higher uncertainty of the future (β = 0.22, 95% CI = 0.17, 0.28) among youth, as well as lower consideration of future consequences (β = −0.15, 95% CI = −0.23, −0.06) among parents. Additionally, severe food insecurity was associated with the lowest future orientation scores for youth and their parents. Our findings indicate that, under conditions of food insecurity, youth and their parents lose sight of the future, which may come at a great cost. Encouraging individuals to engage in thinking about the future or to create future images of their selves may not be meaningful if their basic needs, including access to food, are not met. Programs that provide opportunities to generate income and accumulate assets may have a twofold effect of increasing access to food and improving future orientation.

Comparative analysis of the long-term effect of two families of high-flux polysulfone dialysers on platelet count: a retrospective cross-sectional study.

IntroductionThrombocytopenia is a potential complication of hemodialysis (HD), and its occurrence has been described even with highly biocompatible polysulfone (PSf) membranes. Dialysis units routinely monitor platelet (PLT) count at the beginning of HD sessions. However, considering that the long-term effects on PLT count could easily be missed, the prevalence of HD-related thrombocytopenia could be underestimated. In the present study, we aimed to investigate the following: 1) the long-term impact of HD treatment on PLT count, comparing two families of dialysis membranes made up of similar PSfs; 2) whether the switch between the dialysis membranes studied significantly affects PLT count; and 3) the prevalence and the risk of HD-induced thrombocytopenia according to the dialysis membranes used.MethodsA cross-sectional retrospective study was performed comprising 157 adult chronic HD patients. The HD membranes under investigation were of the series FX, Helixone® Fresenius (Filters A), and Polyflux® Gambro (Filters B). Patients were treated in 4 dialysis units in Southern Switzerland. Data were collected from a centralized computing platform.FindingsPLT count significantly differs between Filters A and B with, respectively, 188 (153–243) ×10E9/L versus 214 (179–255) ×10E9/L (p=0.036). The prevalence of thrombocytopenia was higher for Filter A compared with Filter B (28.4% versus 12.8%; p<0.001). The switch from Filter A to Filter B significantly affected PLT count: from 189 (146–217) ×10E9/L to 217 (163–253) ×10E9/L (p<0.001; analysis on 26 patients). A linear random-intercept model confirmed the results (β coefficient =35.214; SE =5.956; p<0.001). In a mixed-effects logistic regression model, the risk of thrombocytopenia for Filter B was 0.157 (CI =0.056–0.442).DiscussionOur data suggest that among the PSf membranes studied, the FX membrane induced a lasting decrease in PLT count and caused significantly more thrombocytopenia. Prospective studies are warranted to verify our findings.

Omega-3 supplementation from pregnancy to postpartum to prevent depressive symptoms: a randomized placebo-controlled trial

BackgroundLow n-3 polyunsaturated fatty acids (PUFAs) has been linked to depression, but the preventive effect of n-3PUFAs supplementation on maternal depression needs further investigation. We aimed to evaluate the efficacy of a daily dose of n-3 PUFAs supplementation (fish oil) on the prevention of postpartum depression (PPD).MethodsA randomized, placebo-controlled, double blind trial was designed and nested into a cohort study conducted in Rio de Janeiro, Brazil. Sixty pregnant women identified as being at risk for PPD were invited and randomly assigned to receive fish oil capsules [1.8 g (1.08 g of Eicosapentaenoic (EPA) and 0.72 g of Docosapentaenoic (DHA) acids)] or placebo (control). The Edinburgh Postnatal Depression Scale (EPDS) was scored at 5–13 (T0, baseline), 22–24 (T1), 30–32 weeks of gestation (T2) and 4–6 weeks’ postpartum (T3). Supplementation started at week 22–24 of gestation (T1) and lasted for 16 weeks. Serum fatty acids were assayed to evaluate compliance. Prevalence of EPDS ≥11 was the primary outcome, and mean and changes in EPDS score, length of gestation, and birth weight the secondary outcomes. Linear mixed-effect (LME) and random-intercept logistic regression models were performed to test the effect of fish oil supplementation on prevalence of EPDS ≥11 and EPDS scores variation.ResultsIn intention-to-treat (ITT) analysis, at 30–32 weeks’ gestation women in the fish oil presented higher serum concentration of EPA, DHA and lower n-6/n-3 ratio comparing to the control group. There were no differences between intervention and control groups in the prevalence of EPDS ≥11, EPDS scores over time, or in changes in EPDS scores from pregnancy to postpartum in either the ITT or per-protocol analyses. Women in the fish oil group with previous history of depression presented a higher reduction on the EPDS score from the second to the third trimester in the fish oil comparing to the control group in the ITT analyses [−1.0 (−3.0–0.0) vs. -0.0 (−1.0–3.0), P = 0.038). These results were confirmed on the LME model (β = −3.441; 95%CI: -6.532– -0.350, P = 0.029).ConclusionDaily supplementation of 1.8 g of n-3 PUFAs during 16 weeks did not prevent maternal depressive symptoms in a sample of Brazilian women.Trial registrationClinicalTrials.gov Identifier: NCT01660165. Retrospectively registered on 24 May 2012.

Conditional assessment of the impact of a Hausman pretest on confidence intervals

In the analysis of clustered and longitudinal data, which includes a covariate that varies both between and within clusters, a Hausman pretest is commonly used to decide whether subsequent inference is made using the linear random intercept model or the fixed effects model. We assess the effect of this pretest on the coverage probability and expected length of a confidence interval for the slope, conditional on the observed values of the covariate. This assessment has the advantages that it (i) relates to the values of this covariate at hand, (ii) is valid irrespective of how this covariate is generated, (iii) uses exact finite sample results, and (iv) results in an assessment that is determined by the values of this covariate and only two unknown parameters. For two real data sets, our conditional analysis shows that the confidence interval constructed after a Hausman pretest should not be used.

Impact of early palliative care on caregivers of patients with advanced cancer: cluster randomised trial

Early palliative care improves the quality of life (QoL) and satisfaction with care of patients with advanced cancer, but little is known about its effect on caregivers. Here, we report outcomes of caregiver satisfaction with care and QoL from a trial of early palliative care. Twenty-four medical oncology clinics were cluster-randomised, stratified by tumour site (lung, gastrointestinal, genitourinary, breast and gynaecological), to early palliative care team referral, or to standard oncology care with palliative care only as needed. Caregivers of patients with advanced cancer (clinical prognosis of 6-24 months, Eastern Cooperative Oncology Group 0-2) in both trial arms completed validated measures assessing satisfaction with care (FAMCARE-19) and QoL [SF-36v2 Health Survey; Caregiver QoL-Cancer (CQoL-C)], at baseline and monthly for 4 months. We used a multilevel linear random-intercept mixed-effect model to test whether there was improvement in the intervention group relative to the control group over 3 and 4 months. A total of 182 caregivers completed baseline measures (94 intervention, 88 control); 151 caregivers (77 intervention, 74 control) completed at least one follow-up assessment. Satisfaction with care improved in the palliative intervention group compared with controls over 3 months (P = 0.007) and 4 months (P = 0.02). There was no significant improvement in the intervention group compared with controls for CQoL-C (3 months: P = 0.92, 4 months: P = 0.51), Physical Component Summary of the SF-36v2 Health Survey (3 months: P = 0.83, 4 months: P = 0.20), or Mental Component Summary of the SF-36v2 Health Survey (3 months: P = 0.87, 4 months: P = 0.60). Early palliative care increased satisfaction with care in caregivers of patients with advanced cancer. NCT01248624.

Health and growth of Finnish beef calves and the relation to acute phase response

Healthy, thriving calves are essential for beef calf production. We studied the health status and factors associated with the growth of beef calves in six cow-calf herds during the first month of the calves’ lives and at weaning age (200 days).The six herds were visited three times, when calves were approximately 3 days, 16 days and 30 days of age. On each visit calves (n=37) were clinically examined, weighed or measured, blood samples were collected, faecal samples obtained and deep nasopharyngeal swabs were taken. Each blood sample was analysed for acute phase proteins (haptoglobin, serum amyloid-A, fibrinogen), total proteins and albumin, the faecal sample for intestinal tract pathogens (rotavirus, bovine coronavirus, enterotoxigenic Escherichia coli and Salmonella, oocysts of Eimeria coccidia and Cryptosporidium, and nematode eggs), and the nasopharyngeal swab for respiratory tract pathogens (bovine coronavirus (BCV), respiratory syncytial virus (RSV), bacteria and mycoplasma).Clinical diagnosis of respiratory tract disease, diarrhoea or umbilical disease was set at 15.0% for all the three consecutive examinations combined (n=107), but only few pathogens were detected from the samples. The increased levels of acute phase proteins were neither associated with any of the diseases nor with the pathogens. Random intercept linear models were used to explore factors affecting early (3–30 days) and long-term (3–200 days) growth, showing that calves with elevated serum amyloid-A concentrations at the age of 16 days had lower long-term growth. Increased albumin concentration at 30 days of age and higher parity of the dam increased early-term growth. The lack of association between a disease and the acute phase protein may stem from the low disease prevalence in the beef calves examined. The measurement of acute phase proteins of a young calf can help identify animals with possible future growth deficiencies, although the mechanisms through which the association between acute phase proteins and growth has yet to be explained.

Associations between Environmental Quality and Mortality in the Contiguous United States, 2000-2005.

Background:Assessing cumulative effects of the multiple environmental factors influencing mortality remains a challenging task.Objectives:This study aimed to examine the associations between cumulative environmental quality and all-cause and leading cause-specific (heart disease, cancer, and stroke) mortality rates.Methods:We used the overall Environmental Quality Index (EQI) and its five domain indices (air, water, land, built, and sociodemographic) to represent environmental exposure. Associations between the EQI and mortality rates (CDC WONDER) for counties in the contiguous United States (n = 3,109) were investigated using multiple linear regression models and random intercept and random slope hierarchical models. Urbanicity, climate, and a combination of the two were used to explore the spatial patterns in the associations.Results:We found 1 standard deviation increase in the overall EQI (worse environment) was associated with a mean 3.22% (95% CI: 2.80%, 3.64%) increase in all-cause mortality, a 0.54% (95% CI: –0.17%, 1.25%) increase in heart disease mortality, a 2.71% (95% CI: 2.21%, 3.22%) increase in cancer mortality, and a 2.25% (95% CI: 1.11%, 3.39%) increase in stroke mortality. Among the environmental domains, the associations ranged from –1.27% (95% CI: –1.70%, –0.84%) to 3.37% (95% CI: 2.90%, 3.84%) for all-cause mortality, –2.62% (95% CI: –3.52%, –1.73%) to 4.50% (95% CI: 3.73%, 5.27%) for heart disease mortality, –0.88% (95% CI: –2.12%, 0.36%) to 3.72% (95% CI: 2.38%, 5.06%) for stroke mortality, and –0.68% (95% CI: –1.19%, –0.18%) to 3.01% (95% CI: 2.46%, 3.56%) for cancer mortality. Air had the largest associations with all-cause, heart disease, and cancer mortality, whereas the sociodemographic index had the largest association with stroke mortality. Across the urbanicity gradient, no consistent trend was found. Across climate regions, the associations ranged from 2.29% (95% CI: 1.87%, 2.72%) to 5.30% (95% CI: 4.30%, 6.30%) for overall EQI, and larger associations were generally found in dry areas for both overall EQI and domain indices.Conclusions:These results suggest that poor environmental quality, particularly poor air quality, was associated with increased mortality and that associations vary by urbanicity and climate region.Citation:Jian Y, Messer LC, Jagai JS, Rappazzo KM, Gray CL, Grabich SC, Lobdell DT. 2017. Associations between environmental quality and mortality in the contiguous United States, 2000–2005. Environ Health Perspect 125:355–362; http://dx.doi.org/10.1289/EHP119

Gingival Margin Stability After Mucogingival Plastic Surgery. The Effect of Manual Versus Powered Toothbrushing: A Randomized Clinical Trial.

Oscillating-rotating power toothbrushes have been proven to be clinically efficacious. To the best of the authors' knowledge, a clinical evaluation of the safety of these toothbrushes after surgical root coverage procedures has not been published. The aim of this study is to evaluate the gingival margin (GM) stability with the use of an oscillating-rotating toothbrush compared with a manual toothbrush. Sixty healthy individuals with at least one Miller Class I or II gingival recession underwent a surgical root coverage procedure. Soft-bristle manual and powered toothbrushes were given to participants randomly assigned to control and test groups, respectively. Full-mouth plaque score (FMPS), full-mouth bleeding score (FMBS), probing depth (PD), and recession depth (RD) were recorded at baseline and 1, 3, and 6 months after completion of the surgical procedure. Data analyses were performed using linear random-intercept models to take into account within-participant correlations over time. Temporal trend differences across treatments by including treatment-time interaction terms were then tested using a global Wald test. Use of a powered toothbrush resulted in a significantly greater reduction of recorded periodontal clinical indices compared with a manual device (FMPS, P = 0.05; FMBS, P = 0.005; RD, P = 0.004). No significant differences were noticed between the two experimental groups both for PD (P = 0.03) and clinical attachment level (P = 0.11). Complete root coverage was significantly higher in participants who used the powered toothbrush compared with the manual toothbrush at 6 months (control, 66.67%; test, 96.67%; P = 0.002). Use of an oscillating-rotating powered toothbrush with a soft-bristle head resulted in higher GM stability after root coverage procedures compared with the use of a manual soft-bristled toothbrush.

A comparative study on estimation methods to deal with the endogeneity in linear random-intercept models with an extension

ABSTRACTIn this paper, we investigate estimation methods to deal with situations where random intercepts are associated to time-varying covariates in the context of linear mixed models. First, a review of previous ways to deal with this so-called endogeneity issue is presented, then a new method based on shared random effects is proposed. Simulation studies and an empirical example are utilized to assess the performance of our proposed method. It is shown that our new approach is more efficient than most competitors and is robust to the misspecification of the random-effects distributions.

Associations between physical activity, television viewing and postnatal depressive symptoms amongst healthy primiparous mothers

BackgroundPostnatal depression is a serious illness in new mothers. In the general population, physical activity (PA) has been found to reduce the risk of depression, whilst sedentary behaviour (SB; in particular television viewing) has been linked to higher levels of depressive symptoms, yet little is known regarding associations between PA, SB and postnatal depression. This study aimed to investigate associations between PA, television viewing and postnatal depressive symptoms in healthy primiparous mothers.. MethodsCross-sectional survey data were provided by 406 first-time mothers (approximately 3-months postpartum) enrolled in the Melbourne InFANT Extend trial (2012/2013). Women self-reported PA (time spent walking for leisure and transport, and other moderate and vigorous PA), television viewing, and depressive symptoms (CES-D 10). Random intercept linear models examined associations between PA, television viewing and depressive symptoms. ResultsIn crude models total PA was inversely associated with risk of postnatal depressive symptoms (B = −0.122; 95% CI = −0.24, -0.01). In models adjusted for key sociodemographic and behavioural covariates the association did not remain statistically significant. No other associations between PA, television viewing and postnatal depressive symptoms were evident.. ConclusionsPostnatal depressive symptoms may not be related to PA and television viewing in the same way that these behaviours predict depressive symptoms in the general population. Further investigation of the specific domains of PA, as well as different types/contexts of SB and their respective associations with postnatal depressive symptoms is warranted in order to better inform development of targeted interventions aimed at enhancing postnatal mental health..

The impact of health information technology on disparity of process of care.

IntroductionDisparities in the quality of health care and treatment among racial or ethnic groups can result from unequal access to medical care, disparate treatments for similar severities of symptoms, and wide divergence in general health status among individuals. Such disparities may be eliminated through better use of health information technology (IT). Investment in health IT could foster better coordinated care, improve guideline compliance, and reduce the likelihood of redundant testing, thereby encouraging more equitable treatment for underprivileged populations. However, there is little research exploring the impact of health IT investment on disparities of process of care.MethodologyThis study examines the impact of health IT investment on waiting times – from admission to the date of first principle procedure – among different racial and ethnic groups, using patient and hospital data for the state of California collected from 2001 to 2007. The final sample includes 14,056,930 patients admitted with medical diseases to 316 unique, acute-care hospitals over a seven-year period. The linear random intercept and slope model was employed to examine the impacts of health IT investment on waiting time, while controlling for patient, disease, and hospital characteristics.ResultsGreater health IT investment was associated with shorter waiting times, and the reduction in waiting times was greater for non-White than for White patients. This indicates that minority populations could benefit from health IT investment with regard to process of care.ConclusionInvestments in health IT may reduce disparities in process of care.Electronic supplementary materialThe online version of this article (doi:10.1186/s12939-015-0161-3) contains supplementary material, which is available to authorized users.

Examining food purchasing patterns from sales data at a full-service grocery store intervention in a former food desert

BackgroundThe Good Food Junction Grocery Store was opened in a former food desert in the inner city of Saskatoon, Canada. ObjectiveThe purpose of this research was to examine, using grocery store sales data, healthy and less healthful food purchasing over a one-year period beginning eight months after opening by shoppers' neighborhood of residence. DesignA multilevel cross sectional design was used. The sample consisted of members of the Good Food Junction with a valid address in Saskatoon, Saskatchewan. All purchases made by members who reported their postal code of residence from May 15, 2013 to April 30, 2014 were analyzed. The outcome variable was the total amount spent on foods in 11 food groups. Linear random intercept models with three levels were fit to the data. ResultsShoppers who were residents of former food desert neighborhoods spent $0.7 (95% CI: 0.2 to 1.2) more on vegetables, and $1.2 (95% CI: −1.8 to −0.6) less on meat, and $1.1 (95% CI: −2.0 to −0.3) less on prepared foods than shoppers who did not reside in those neighborhoods. ConclusionsWhen given geographical access to healthy food, people living in disadvantaged former food desert neighborhoods will take advantage of that access.

Age at migration and self-rated health trajectories after age 50: understanding the older immigrant health paradox.

This research contributes to the "immigrant health paradox" debate by testing the hypothesis that older age at migration is associated with the increased risk of poor health in later life. Using the 1992-2008 Health and Retirement Study, I construct linear random-intercept models to estimate self-rated health (SRH) trajectories after age 50 for the native and foreign born by age at migration. At age 50, both Hispanic and non-Hispanic foreign born report better SRH compared with their native-born counterparts, net of race, gender, and education. Non-Hispanic foreign born who migrated after age 35 and Hispanic foreign born who migrated after age 18, however, experience steeper decline in SRH after age 50, which results in a health disadvantage vis-à-vis the native born in old age. Education has a smaller protective effect on SRH for the foreign born, especially those who migrated as adults. Age at migration is an important factor for understanding health status of older immigrants. Steeper health decline in later life of the foreign born who migrated in advanced ages may be related to longer exposure to unfavorable conditions in home countries and limited opportunities for incorporation in the United States.

Can valproic acid be an inducer of clozapine metabolism?

Prior clozapine studies indicated no effects, mild inhibition or induction of valproic acid (VPA) on clozapine metabolism. The hypotheses that (i) VPA is a net inducer of clozapine metabolism, and (ii) smoking modifies this inductive effect were tested in a therapeutic drug monitoring study. After excluding strong inhibitors and inducers, 353 steady-state total clozapine (clozapine plus norclozapine) concentrations provided by 151 patients were analyzed using a random intercept linear model. VPA appeared to be an inducer of clozapine metabolism since total plasma clozapine concentrations in subjects taking VPA were significantly lower (27% lower; 95% confidence interval, 14-39%) after controlling for confounding variables including smoking (35% lower, 28-56%). Prospective studies are needed to definitively establish that VPA may (i) be an inducer of clozapine metabolism when induction prevails over competitive inhibition, and (ii) be an inducer even in smokers who are under the influence of smoking inductive effects on clozapine metabolism.

Occlusal tooth wear in the general population of Germany: Effects of age, sex, and location of teeth

Tooth wear is an increasing problem in a society where people are living longer. The purpose of this study was to assess the effect of age, sex, and location of teeth on the severity of tooth wear andto determine the prevalence of dentin exposure in the general population of Germany. Tooth wear was measured in casts of both jaws of 836 persons with a 6-point (0-5) ordinal ratingscale. Linear random-intercept regression models with the covariates of age, sex, jaw, and tooth group (with the participant as a grouping variable) were computed to determine the association of these covariates with tooth wear of a single tooth. The mean tooth wear score across all age groups, both sexes, and all teeth was 2.9 (standard deviation, 0.8), and theprevalence of teeth with exposed dentin was 23.4%. The participants' age was correlated with the mean tooth wear scores(r=0.51). The tooth wear level among women was on average 0.15 units lower than among men, and tooth wear wason average 0.59 units higher for anterior teeth than for posterior teeth. Increased tooth wear in anterior teeth may be due to the initially predominant guidance by anterior teeth, withage-related linear progress in tooth wear. Occlusal tooth wear scores and dentin exposure increase with age.

The randomized shortened dental arch study: temporomandibular disorder pain

The aim of this multi-center, randomized controlled trial was to assess the impact of missing posterior support on the risk for temporomandibular disorder (TMD) pain by comparing patients with either shortened dental arches (SDA) or molar replacement by removable dental prostheses (RDP). A sample of 215 patients with bilateral molar loss in at least one jaw was consecutively recruited in 14 prosthodontic departments of dental schools in Germany. Of the initial sample, 152 patients (mean age: 59.7years; 53.9% female) received randomly allocated interventions (SDA: n = 71; RDP: n = 81). Presence of TMD pain was assessed using patients' self-reports and was verified by physical examination and by pain intensity, as the mean of current pain, worst pain, and average pain in the last 6 months, with 10-point ordinal rating scales. Assessments were performed before treatment and at follow-ups until 60 months after treatment. Impact of interventions on TMD risk and pain intensity was computed by applying logistic and linear random-intercept models. Tooth replacement (RDP) did not significantly change the risk for self-reported (odds ratio [OR]: 1.1; confidence interval [CI]: 0.4 to 3.4) or clinically verified (OR: 0.7; CI: 0.1 to 4.3) TMD pain compared to no tooth replacement (SDA). Mean characteristic pain intensity was virtually identical in both groups (Coeff: 0.01; CI: -0.30 to 0.32). Retaining or preservation of an SDA is not a major risk factor for TMD pain over the course of 5 years when compared to molar replacement with RPDs. Seemingly, missing molars do not have to be replaced in order to prevent TMD pain.

Predictors of disordered eating in adolescence and young adulthood

We investigated longitudinal predictors for disordered eating from early adolescence to young adulthood (12–34 years) across gender and different developmental phases among Norwegian young people. Survey data from a population-based sample were collected at four time points (T) over a 13-year time span. A population-based sample of 5,679 females and males at T1 and T2, 2,745 at T3 and 2,718 at T4 were included in analyses, and linear regression and random intercept models were applied. In adolescence, initial disordered eating and parental overprotectiveness were more strongly related to disordered eating among females, whereas loneliness was a stronger predictor for adolescent males. Initial disordered eating during early adolescence predicted later disordered eating more strongly in late- than mid-adolescence. In young adulthood, no significant gender-specific risk factors were found. The findings provide support for both shared and specific risk factors for the developmental psychopathology of disordered eating.

“Push as Hard as You Can” Instruction for Telephone Cardiopulmonary Resuscitation: A Randomized Simulation Study

BackgroundThe medical priority dispatch system (MPDS®) assists lay rescuers in protocol-driven telephone-assisted cardiopulmonary resuscitation (CPR). ObjectiveOur aim was to clarify which CPR instruction leads to sufficient compression depth. MethodsThis was an investigator-blinded, randomized, parallel group, simulation study to investigate 10 min of chest compressions after the instruction “push down firmly 5 cm” vs. “push as hard as you can.” Primary outcome was defined as compression depth. Secondary outcomes were participants exertion measured by Borg scale, provider's systolic and diastolic blood pressure, and quality values measured by the skill-reporting program of the Resusci® Anne Simulator manikin. For the analysis of the primary outcome, we used a linear random intercept model to allow for the repeated measurements with the intervention as a covariate. ResultsThirteen participants were allocated to control and intervention. One participant (intervention) dropped out after min 7 because of exhaustion. Primary outcome showed a mean compression depth of 44.1 mm, with an inter-individual standard deviation (SDb) of 13.0 mm and an intra-individual standard deviation (SDw) of 6.7 mm for the control group vs. 46.1 mm and a SDb of 9.0 mm and SDw of 10.3 mm for the intervention group (difference: 1.9; 95% confidence interval −6.9 to 10.8; p = 0.66). Secondary outcomes showed no difference for exhaustion and CPR-quality values. ConclusionsThere is no difference in compression depth, quality of CPR, or physical strain on lay rescuers using the initial instruction “push as hard as you can” vs. the standard MPDS® instruction “push down firmly 5 cm.”

Investigation of Ways to Handle Sampling Weights for Multilevel Model Analyses

When analysts estimate a multilevel model using survey data, they often use weighted procedures using multilevel sampling weights to correct the effect of unequal probabilities of selection. This study addresses the impacts of including sampling weights and the consequences of ignoring them by assessing the performance of four approaches: the multilevel pseudo–maximum likelihood (MPML), the probability-weighted iterative generalized least squares (PWIGLS), the naive (ignoring sampling weights), and the sample distribution methods for a linear random-intercept model under a two-stage clustering sampling design. When inclusion probabilities are correlated with the values of outcome variable conditioning on the model covariates, the sampling design becomes informative. The results show that whether a sampling design is informative and at which stage of the sampling design it is informative have substantial impacts on the estimation. The results also show that the level of variation of sampling weights is correlated with the bias of estimates. A higher level of variation of sampling weights is associated with a higher level of bias when a sampling design is informative; however, under a noninformative design, the level of variation of sampling weights may not necessarily associate with biased results. Ignoring an informative sampling design at the first stage will result in biased estimates on the intercept and variance of random effect, whereas ignoring an informative sampling design at the second stage will lead to slightly underestimated fixed effects and residual variance, in addition to the biased estimates on the intercept and variance of random effect. Including the sampling weights as the hybrid methods (MPML and PWIGLS) may still produce biased estimates on the intercept and variance of random effect and slightly underestimated fixed effects and residual variance. The sample distribution method may give unbiased estimates, but it depends on the correct specification of the sampling process.

The Role of Sample Cluster Means in Multilevel Models

The paper explores some issues related to endogeneity in multilevel models, focusing on the case where the random effects are correlated with a level 1 covariate in a linear random intercept model. We consider two basic specifications, without and with the sample cluster mean. It is generally acknowledged that the omission of the cluster mean may cause omitted-variable bias. However, it is often neglected that the inclusion of the sample cluster mean in place of the population cluster mean entails a measurement error that yields biased estimators for both the slopes and the variance components. In particular, the contextual effect is attenuated, while the level 2 variance is inflated. We derive explicit formulae for measurement error biases that allow us to implement simple post-estimation corrections based on the reliability of the covariate. In the first part of the paper, the issue is tackled in a standard framework where the population cluster mean is treated as a latent variable. Later we consider a different framework arising when sampling from clusters of finite size, where the latent variable methods may have a poor performance, and we show how to effectively modify the measurement error correction. The theoretical analysis is supplemented with a simulation study and a discussion of the implications for effectiveness evaluation.