Background Hospital-acquired anemia (HAA), which negatively affects morbidity, mortality, and resource use, is a common iatrogenic effect of hospitalization, especially in intensive care units. Contributing factors include phlebotomy, intravenous fluid resuscitation, and medications. Objective To determine the prevalence of HAA among adult patients in a southwest Florida intensive care unit and evaluate the relationship between decline in hemoglobin level and length of intensive care unit stay. Methods A retrospective, exploratory medical record review of a 6-month convenience sample (N = 151) was conducted. Hospital-acquired anemia was defined as a 10% or greater decline in hemoglobin level from admission to discharge. Results Of 108 patients meeting inclusion criteria, 69 (64%) developed HAA. Stays were significantly longer for the 69 patients with HAA than for the 39 patients without HAA (mean [SD], 5.9 [5.8] days vs 3.3 [3.6] days; P = .005). The mean (SD) decline in hemoglobin level was 20% (8%) for patients with HAA and 4% (2%) for patients without HAA (P < .001). A sensitivity analysis excluding 6 outliers (n = 102) confirmed that stays were significantly longer for patients with HAA than for patients without HAA (mean [SD], 5.1 [4.2] days vs 2.8 [1.7] days; P < .001). Conclusion Hospital-acquired anemia is prevalent in this intensive care unit and correlates with longer stays. Although causality regarding phlebotomy was not directly measured, the high prevalence of HAA supports implementing blood conservation strategies to improve clinical outcomes and reduce iatrogenic blood loss.
Background Pulse oximetry is an essential, noninvasive tool in critical care, yet mounting evidence indicates that readings may be inaccurate in patients with darker skin pigmentation. These discrepancies can delay recognition of hypoxemia and contribute to racial disparities in clinical outcomes. Objective To quantify the difference between oxygen saturation as measured by pulse oximetry and arterial blood gas measurements across a spectrum of skin pigmentation using objective measures of melanin and skin tone. Methods A prospective observational study was conducted in an adult critical care unit. Skin pigmentation was measured using reflectance spectrophotometry. Simultaneous pulse oximetry and arterial blood gas readings were collected during routine arterial blood gas sampling. Bias (pulse oximetry value – arterial blood gas value) was calculated for each measurement pair. Multivariate linear regression models were used to assess the association between skin pigmentation and measurement bias, adjusting for potential confounders such as age, sex, and perfusion index. Results Forty-two measurement pairs were analyzed. A weak and nonsignificant relationship was found between the individual typology angle and melanin index and the difference between pulse oximetry and arterial blood gas measures. Conclusions Individual typology angle and melanin index are not strong predictors of the difference in oxygen readings between pulse oximeters and arterial blood gas measurements, even in critically ill patients. The weak fit of all models and the lack of statistical significance indicate that other unmeasured factors may play a more substantial role in explaining the variability.
Background Dysbiotic gut microbiomes are linked to morbidity, mortality, and poor outcomes, so it is critical to understand their influencing factors. Objectives To investigate the impact of nosocomial and socioecological factors on the gut microbiome in newborns with congenital heart disease (CHD). Methods This was a secondary analysis of a prospective cohort of newborns with CHD matched to healthy newborns by age, sex, and race and ethnicity. Data on feeding type and routes and disease-related factors were collected from electronic medical records. Stool samples were collected at 3 perioperative time points. Gut microbiome diversity and abundance were analyzed using QIIME 2. Results Newborns with CHD (n = 41) showed higher alpha diversity (P = .007) and dissimilar beta diversity (P = .001) versus controls (n = 40). Beta diversity was dissimilar between those receiving human milk versus formula (P = .05), between those receiving fortified versus nonfortified human milk (P = .04), and between those receiving naso-orogastric tube feeds versus other routes (P = .02). Newborns with CHD fed only human milk had higher relative abundance of Haemophilus (P = .04), those with longer duration of aortic cross-clamping had lower relative abundance of Streptococcus (P = .10), and those of higher socioeconomic deprivation had higher relative abundance of Enterococcus (P = .03). Conclusions Feeding tubes, formula exposure, aortic cross-clamp, and socioeconomic status were associated with altered gut microbiome diversity. Further study is needed to determine the causal relationships leading to gut microbial dysbiosis in newborns in the cardiac intensive care unit.
