RESUMEN
Missing data is a common problem in medical research, and is commonly addressed using multiple imputation. Although traditional imputation methods allow for valid statistical inference when data are missing at random (MAR), their implementation is problematic when the presence of missingness depends on unobserved variables, that is, the data are missing not at random (MNAR). Unfortunately, this MNAR situation is rather common, in observational studies, registries and other sources of real-world data. While several imputation methods have been proposed for addressing individual studies when data are MNAR, their application and validity in large datasets with multilevel structure remains unclear. We therefore explored the consequence of MNAR data in hierarchical data in-depth, and proposed a novel multilevel imputation method for common missing patterns in clustered datasets. This method is based on the principles of Heckman selection models and adopts a two-stage meta-analysis approach to impute binary and continuous variables that may be outcomes or predictors and that are systematically or sporadically missing. After evaluating the proposed imputation model in simulated scenarios, we illustrate it use in a cross-sectional community survey to estimate the prevalence of malaria parasitemia in children aged 2-10 years in five regions in Uganda.
Asunto(s)
Investigación Biomédica , Niño , Humanos , Estudios Transversales , Uganda/epidemiologíaRESUMEN
BACKGROUND: Diffuse alveolar hemorrhage (DAH) occurs during the course of autoimmune disease and may be life threatening. The objective was to assess characteristics and prognosis factors of DAH who required intensive care unit (ICU) admission in patients with autoimmune diseases. METHODS: French multicenter retrospective study including patients presenting DAH related to autoimmune diseases requiring ICU admission from 2000 to 2016. RESULTS: One hundred four patients (54% of men) with median age of 56 [32-68] years were included with 79 (76%) systemic vasculitis and 25 (24%) connective tissue disorders. All patients received steroids, and 72 (69%), 12 (11.5%), and 57 (55%) patients had cyclophosphamide, rituximab, and plasma exchanges, respectively. During ICU stay, 52 (50%), 36 (35%), and 55 (53%) patients required mechanical ventilation, vasopressor use, and renal replacement therapy, respectively. Factors associated with mechanical ventilation weaning were age (HR [95%CI] 0.97 [0.96-0.99] per 10 years, p < 0.0001), vasculitis-related DAH (0.52 [0.27-0.98], p = 0.04), and time from dyspnea onset to ICU admission (0.99 [0.99-1] per day, p = 0.03). ICU mortality was 15%. Factors associated with alive status at ICU discharge were chronic cardiac failure (HR [95%CI] 0.37 [0.15-0.94], p = 0.04), antiphospholipid syndrome-related DAH (3.17 [1.89-5.32], p < 0.0001), SAPS II (0.98 [0.97-0.99], p = 0.007), and oxygen flow at ICU admission (0.95 [0.91-0.99] per liter/min, p = 0.04). CONCLUSION: DAH in autoimmune diseases is a life-threatening complication which requires mechanical ventilation in half of the cases admitted to ICU.
Asunto(s)
Enfermedades Autoinmunes/complicaciones , Hemorragia/etiología , Alveolos Pulmonares/anomalías , Adulto , Anciano , Enfermedades Autoinmunes/fisiopatología , Femenino , Francia , Hemorragia/fisiopatología , Humanos , Unidades de Cuidados Intensivos/organización & administración , Unidades de Cuidados Intensivos/estadística & datos numéricos , Masculino , Persona de Mediana Edad , Alveolos Pulmonares/fisiopatología , Estudios RetrospectivosRESUMEN
Introduction: Burn injury is associated with a high risk of death. Whether a pattern of immune and inflammatory responses after burn is associated with outcome is unknown. The aim of this study was to explore the association between systemic immune and inflammatory responses and outcome in severely-ill burn patients. Materials and Methods: Innate immunity, adaptive immunity, activation and stress and inflammation biomarkers were collected at admission and days 2, 7, 14, and 28 in severely-ill adult burn patients. Primary endpoint was mortality at day 90, secondary endpoint was secondary infections. Healthy donors (HD) served as controls. Multiple Factorial Analysis (MFA) was used to identify patterns of immune response. Results: 50 patients were included. Age was 49.2 (44.2-54.2) years, total burn body surface area was 38.0% (32.7-43.3). Burn injury showed an upregulation of adaptive immunity and activation biomarkers and a down regulation of innate immunity and stress/inflammation biomarkers. High interleukin-10 (IL-10) at admission was associated with risk of death. However, no cluster of immune/inflammatory biomarkers at early timepoints was associated with mortality. HLA-DR molecules on monocytes at admission were associated with bacterial infections and septic shock. Later altered immune/inflammatory responses in patients who died may had been driven by the development of septic shock. Conclusion: Burn injury induced an early and profound upregulation of adaptive immunity and activation biomarkers and a down regulation of innate immunity and stress/inflammation biomarkers. Immune and inflammatory responses were associated with bacterial infection and septic shock. Absence of immune recovery patterns was associated with poor prognosis.