Multiple Imputation for Missing Laboratory Data: An Example from Infectious Disease Epidemiology

Mulla, Zuber D.; Seo, Byungtae; Nuwayhid, B.

doi:10.1016/j.annepidem.2009.08.002

Cited by 14 publications

(12 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Attrition at one-year follow-up was high given the hard-to-reach population under study with a total of 284 participants completing the assessment (48% completed one-year follow-up). However, theoretically, large amounts of missing data should not be a critical problem as long as the sample size is large enough and the missing values are missing at random (Graham, 2009; Mulla et al, 2009). The analytic sample excluded those who indicated condom use was not applicable (n = 71) or had no sex in the past four months (n = 13).…”

Section: Methodsmentioning

confidence: 99%

Predictive utility and measurement properties of the Strength of Implementation Intentions Scale (SIIS) for condom use

Nydegger

Ames

Stacy

2017

Social Science & Medicine

View full text Add to dashboard Cite

Background Implementation intentions are situation-linked action plans that increase health behaviors such as condom use. Few studies have measured the strength of implementation intentions, especially regarding condom use. Non-injection drug users are at high risk for HIV due to risky sexual practices. Substance use before sex may increase risky sexual behaviors, and implementation intentions are a promising, practical route for HIV prevention. Purpose This prospective study assessed the measurement properties of the newly developed Strength of Implementation Intentions Scale (SIIS) for condom use and investigated whether the SIIS predicted condom use one-year later. The study evaluated whether behavioral intentions and alcohol or methamphetamine use before sex modified the relationship between the SIIS and condom use. Methods Participants were 590 non-injection drug users convicted of nonviolent drug offenses in court-mandated outpatient drug diversion programs. They completed self-directed computer assessments at baseline and at one-year follow-up. Measures included demographics, behavioral intentions to use condoms, the SIIS for condom use, alcohol and methamphetamine use before or during sex, and condom use. Psychometric properties, direct, and interaction effects of the SIIS were evaluated using structural equation modeling and general linear regressions. Results The SIIS, behavioral intentions, and condom use were positively correlated and discriminant validity between all three constructs; the SIIS predicted condom use one-year later in this high-risk population. No significant interactions were detected. Conclusions The SIIS is predictive of condom use behavior, is psychometrically sound and distinct from behavioral intentions, and can be incorporated into HIV prevention interventions and used as a mediator variable. In addition, it can be used as an outcome measure where a follow-up is not feasible. The SIIS can also be tailored for additional HIV-related behaviors such as HIV-testing, adhering to antiretroviral medication, prescribing and adhering to pre-exposure prophylaxis, and other health behaviors.

show abstract

Section: Methodsmentioning

confidence: 99%

Predictive utility and measurement properties of the Strength of Implementation Intentions Scale (SIIS) for condom use

Nydegger

Ames

Stacy

2017

Social Science & Medicine

View full text Add to dashboard Cite

show abstract

“…Hence, in EMR systems we might find cholesterol levels or blood pressure measurements for some patients but not all. The presence of such information may be related to the patient's disease state; consequently, missing values cannot be considered random 41 , 42 …”

Section: Methodological Challengesmentioning

confidence: 99%

“…The presence of such information may be related to the patient's disease state; consequently, missing values cannot be considered random. 41,42 soMe solutions…”

Section: Data Issuesmentioning

confidence: 99%

Assessing the Comparative Effectiveness of Newly Marketed Medications: Methodological Challenges and Implications for Drug Development

Schneeweiß

Gagne

Glynn

et al. 2011

Clin Pharmacol Ther

166

160

View full text Add to dashboard Cite

Comparative-effectiveness research (CER) aims to produce actionable evidence regarding the effectiveness and safety of medical products and interventions as they are used outside of controlled research settings. Although CER evidence regarding medications is particularly needed shortly after market approval, key methodological challenges include (i) potential bias due to channeling of patients to the newly marketed medication because of various patient-, physician-, and system-related factors; (ii) rapid changes in the characteristics of the user population during the early phase of marketing; and (iii) lack of timely data and the often small number of users in the first few months of marketing. We propose a mix of approaches to generate comparative-effectiveness data in the early marketing period, including sequential cohort monitoring with secondary health-care data and propensity score (PS) balancing, as well as extended follow-up of phase III and phase IV trials, indirect comparisons of placebo-controlled trials, and modeling and simulation of virtual trials.

show abstract

“…All cases had subtype information available in gag , env or both gene regions, but among controls, 43/501 (8.6%) were missing all subtype data, including 34/332 (10.2%) from eastern African and 9/169 (5.3%) from southern Africa, due to low HIV-1 plasma viral loads preventing adequate viral amplification. To avoid bias because of control exclusion due to missing subtype data, we performed multiple imputation with 20 datasets imputed using Markov chain Monte Carlo methods 26 .…”

Section: Methodsmentioning

confidence: 99%

HIV-1 subtype C is not associated with higher risk of heterosexual HIV-1 transmission

Kahle¹,

Campbell²,

Lingappa

et al. 2014

Aids

View full text Add to dashboard Cite

Background HIV-1 subtype C has emerged as the most prevalent strain of HIV-1 worldwide, leading to speculation that subtype C may be more transmissible than other subtypes. We compared the risk of HIV-1 transmission for subtype C versus non-C subtypes (A, D, G and recombinant forms) among heterosexual African HIV-1 serodiscordant couples. Methods We conducted a nested case-control analysis using data from two prospective cohort studies of heterosexual HIV-1 serodiscordant couples from 6 countries in eastern and southern Africa. Cases (N=121) included incident HIV-1 transmissions that were established as linked within the serodiscordant partnership by viral sequencing; controls (N=501) were non-transmitting HIV-1 infected partners. Subtype was determined for partial env and gag genes. Multiple logistic regression controlled for age and gender of the HIV-1 infected partner and self-reported unprotected sex. Plasma and genital HIV-1 RNA concentrations were compared between subtype C and non-C subtypes using generalized estimating equations. Results HIV-1 subtype C was not associated with increased risk of HIV-1 transmission compared to non-C subtypes: env adjusted odds ratio (adjOR) 1.14 (95% confidence interval [CI] 0.74–1.75, p=0.6) and gag adjOR 0.98 (95% CI 0.63–1.52, p=0.9). Plasma and genital HIV-1 RNA levels did not differ significantly for subtype C versus non-C. Conclusion In a geographically diverse population of heterosexual African HIV-1 serodiscordant couples, subtype C was not associated with greater risk of HIV-1 transmission compared to non-C subtypes, arguing against the hypothesis that subtype C is more transmissible compared to other common subtypes.

show abstract

Multiple Imputation for Missing Laboratory Data: An Example from Infectious Disease Epidemiology

Cited by 14 publications

References 10 publications

Predictive utility and measurement properties of the Strength of Implementation Intentions Scale (SIIS) for condom use

Predictive utility and measurement properties of the Strength of Implementation Intentions Scale (SIIS) for condom use

Assessing the Comparative Effectiveness of Newly Marketed Medications: Methodological Challenges and Implications for Drug Development

HIV-1 subtype C is not associated with higher risk of heterosexual HIV-1 transmission

Contact Info

Product

Resources

About