A simple method for estimating relative risk using logistic regression

Grupo Latinoamericano de Investigaciones Epidemiológicas, Organización Latinoamericana para el Fomento de la Investigación en Salud, Bucaramanga, Colombia.
BMC Medical Research Methodology (Impact Factor: 2.27). 02/2012; 12(1):14. DOI: 10.1186/1471-2288-12-14
Source: PubMed


Odds ratios (OR) significantly overestimate associations between risk factors and common outcomes. The estimation of relative risks (RR) or prevalence ratios (PR) has represented a statistical challenge in multivariate analysis and, furthermore, some researchers do not have access to the available methods. Objective: To propose and evaluate a new method for estimating RR and PR by logistic regression.
A provisional database was designed in which events were duplicated but identified as non-events. After, a logistic regression was performed and effect measures were calculated, which were considered RR estimations. This method was compared with binomial regression, Cox regression with robust variance and ordinary logistic regression in analyses with three outcomes of different frequencies.
ORs estimated by ordinary logistic regression progressively overestimated RRs as the outcome frequency increased. RRs estimated by Cox regression and the method proposed in this article were similar to those estimated by binomial regression for every outcome. However, confidence intervals were wider with the proposed method.
This simple tool could be useful for calculating the effect of risk factors and the impact of health interventions in developing countries when other statistical strategies are not available.

Download full-text


Available from: Fredi Diaz-Quijano, Dec 13, 2013
    • "Calculation of multivariable odds ratio was conducted by including variables with a p < 0.2 in the univariable analysis of the multivariable model. Age-and sex-adjusted odds ratio (aOR) and corrected risk ratio (cRR) (Diaz-Quijano, 2012) were also estimated. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We initiated a survey to collect basic data on the frequency and regional distribution of various zoonoses in 722 employees of forestry enterprises in the German state of North Rhine-Westphalia (NRW) from 2011 to 2013. Exposures associated with seropositivity were identified to give insight into the possible risk factors for infection with each pathogen. 41.2% of participants were found to be seropositive for anti-Bartonella IgG, 30.6% for anti-Borrelia burgdorferi IgG, 14.2% for anti-Leptospira IgG, 6.5% for anti-Coxiella burnetii IgG, 6.0% for anti-Hantavirus IgG, 4.0% for anti-Francisella tularensis IgG, 3.4% for anti-TBE-virus IgG, 1.7% for anti-Echinococcus IgG, 0.0% for anti-Brucella IgG and anti-XMRV IgG. Participants seropositive for B. burgdorferi were 3.96 times more likely to be professional forestry workers (univariable analysis: OR 3.96; 95% CI 2.60-6.04; p<0.001); and participants seropositive for Hantavirus 3.72 times more likely (univariable analysis: OR 3.72; 95% CI 1.44-9.57; p=0.007). This study found a surprisingly high percentage of participants seropositive for anti-B. henselae IgG and for anti-F. tularensis IgG. The relatively high seroprevalence for anti-Leptospira IgG seen in this study could be related to living conditions rather than to exposure at work. No specific risk for exposure to C. burnetii and Echinococcus was identified, indicating that neither forestry workers nor office workers represent a risk population and that NRW is not a typical endemic area. Forestry workers appear to have higher risk for contact with B. burgdorferi-infected ticks and a regionally diverse risk for acquiring Hantavirus-infection. The regional epidemiology of zoonoses is without question of great importance for public health. Knowledge of the regional risk factors facilitates the development of efficient prevention strategies and the implementation of such prevention measures in a sustainable manner.
    No preview · Article · Aug 2015 · International journal of medical microbiology: IJMM
  • Source
    • "New insights into inflammation in atherosclerosis have practical clinical applications in risk stratification.18)19)20) Pentraxin 3 (PTX3) was initially described as an early marker for primary local activation of innate immunity and inflammatory response.21)22)23) Because PTX3 release is likely to be a specific response to vascular damage, PTX3 levels may provide more explicit information on development and progression of atherosclerosis than nonspecific markers like C-reactive protein.24)25)26)27) "
    [Show abstract] [Hide abstract]
    ABSTRACT: Background and Objectives The aim of this study was to examine the hypothesis that pentraxin 3 (PTX3) can have a diagnostic value for predicting anatomical complexity of coronary artery stenosis as measured by the Synergy between PCI with Taxus and Cardiac Surgery (SYNTAX) score. Subjects and Methods We investigated the association of systemic arterial PTX3 with SYNTAX score among 500 patients with ischemic heart disease assigned to medical treatment (251), percutaneous coronary intervention (PCI) (197), or coronary artery bypass graft (CABG) (52). Results The clinical judgment of the cardiologists was near-perfectly concordant with the SYNTAX score. Mean {99% confidence intervals (CIs)} SYNTAX scores were 5.8 (5.1-6.6), 18.4 (17.1-19.8), and 33.2 (32.8-33.6) in patients assigned to medical therapy, PCI, and CABG, respectively. The AROC (95% CIs) for discriminating between patients with and without a high SYNTAX score (>23) was 0.920 (0.895-0.946) for systemic arterial levels of PTX3. As the systemic arterial level of PTX3 increased, the SYNTAX scores also increased almost in a curvilinear fashion, with the value corresponding to the SYNTAX score of 23 being 0.29 ng · dL-1. This cutpoint achieved a sensitivity of 0.66 (0.57-0.74), a specificity of 0.94 (0.91-0.96), a positive predictive value of 0.79 (0.70-0.87), and a negative predictive value of 0.89 (0.85-0.92). Conclusion We observed that systemic arterial levels of PTX3 were associated with the SYNTAX score in a curvilinear fashion. The discriminatory power of systemic arterial levels of PTX3 for a high SYNTAX score was excellent. The interesting finding of this study was the near perfect concordance between the decisions made by the cardiologists based on their clinical judgment and the SYNTAX score. The systemic arterial PTX3 level of 0.29 ng · dL-1 was highly specific for diagnosing complex coronary artery stenosis.
    Full-text · Article · Jul 2014 · Korean Circulation Journal
  • Source
    • "The LR on the modified dataset will be mathematically equivalent to log binomial model but uses the logit link instead of log therefore it ensures that prediction be between 0 and 1. Also, this method directly provides RR instead of OR and avoids the convergence problem of log binomial model [5]. This form of analysis over represents the total cases. "

    Full-text · Dataset · Jan 2014
Show more