Race, Ethnicity, Language, and the Treatment of Low-Risk Febrile Infants

Importance

Febrile infants at low risk of invasive bacterial infections are unlikely to benefit from lumbar puncture, antibiotics, or hospitalization, yet these are commonly performed. It is not known if there are differences in management by race, ethnicity, or language.

Objective

To investigate associations between race, ethnicity, and language and additional interventions (lumbar puncture, empirical antibiotics, and hospitalization) in well-appearing febrile infants at low risk of invasive bacterial infection.

Design, Setting, and Participants

This was a multicenter retrospective cross-sectional analysis of infants receiving emergency department care between January 1, 2018, and December 31, 2019. Data were analyzed from December 2022 to July 2023. Pediatric emergency departments were determined through the Pediatric Emergency Medicine Collaborative Research Committee. Well-appearing febrile infants aged 29 to 60 days at low risk of invasive bacterial infection based on blood and urine testing were included. Data were available for 9847 infants, and 4042 were included following exclusions for ill appearance, medical history, and diagnosis of a focal infectious source.

Exposures

Infant race and ethnicity (non-Hispanic Black, Hispanic, non-Hispanic White, and other race or ethnicity) and language used for medical care (English and language other than English).

Main Outcomes and Measures

The primary outcome was receipt of at least 1 of lumbar puncture, empirical antibiotics, or hospitalization. We performed bivariate and multivariable logistic regression with sum contrasts for comparisons. Individual components were assessed as secondary outcomes.

Results

Across 34 sites, 4042 infants (median [IQR] age, 45 [38-53] days; 1561 [44.4% of the 3516 without missing sex] female; 612 [15.1%] non-Hispanic Black, 1054 [26.1%] Hispanic, 1741 [43.1%] non-Hispanic White, and 352 [9.1%] other race or ethnicity; 3555 [88.0%] English and 463 [12.0%] language other than English) met inclusion criteria. The primary outcome occurred in 969 infants (24%). Race and ethnicity were not associated with the primary composite outcome. Compared to the grand mean, infants of families that use a language other than English had higher odds of the primary outcome (adjusted odds ratio [aOR]; 1.16; 95% CI, 1.01-1.33). In secondary analyses, Hispanic infants, compared to the grand mean, had lower odds of hospital admission (aOR, 0.76; 95% CI, 0.63-0.93). Compared to the grand mean, infants of families that use a language other than English had higher odds of hospital admission (aOR, 1.08; 95% CI, 1.08-1.46).

Conclusions and Relevance

Among low-risk febrile infants, language used for medical care was associated with the use of at least 1 nonindicated intervention, but race and ethnicity were not. Secondary analyses highlight the complex intersectionality of race, ethnicity, language, and health inequity. As inequitable care may be influenced by communication barriers, new guidelines that emphasize patient-centered communication may create disparities if not implemented with specific attention to equity.

Listening Beyond Auscultating: A Quality Initiative to Improve Communication Scores in the Hospital Consumer Assessment of Health Care Practitioners and Systems Survey

INTRODUCTION

Physician communication is critical to patient care. However, integration of sound communication practice with clinical workflows has proven difficult. In this quality improvement initiative, medical students used the rapid improvement model to test interventions that could enhance patients' perception of listening by physicians as measured by the Hospital Consumer Assessment of Healthcare Providers and Systems survey.

METHODS

Literature review and process analysis yielded 42 potential interventions, of which 24 were feasible for implementation. Small-scale testing established the 4 most promising interventions; pilot testing was subsequently undertaken on the entire Medicine service. Patient and physician feedback guided further refinement. The final intervention used a structured reminder embedded in the electronic health record to direct physicians to begin interviews by eliciting patient concerns.

RESULTS

Patient concerns elicited after implementation included pain symptoms (28%), disease or treatment course (16%), and discharge planning (10%). In the Hospital Consumer Assessment of Healthcare Providers and Systems survey, physician listening scores rose from a 2014 average of 73.6% to 77% in 2015.

DISCUSSION

Among 24 tested interventions, an open-ended question was most feasible and had the greatest perceived impact by hospitalists and patients. A structured reminder embedded in required electronic medical record documentation facilitated the behavioral change without being overly burdensome to physicians and established a mechanism to enact change in practice.

CONCLUSION

Medical students used established improvement methods to promote patient-centered care and align patient and physician agendas, providing a strategy to improve hospitalized patients' perceptions of physician listening.

Trends in the supply of California's emergency departments and inpatient services, 2005-2014: a retrospective analysis

OBJECTIVES

Given increasing demand for emergency care, there is growing concern over the availability of emergency department (ED) and inpatient resources. Existing studies of ED bed supply are dated and often overlook hospital capacity beyond ED settings. We described recent statewide trends in the capacity of ED and inpatient hospital services from 2005 to 2014.

DESIGN

Retrospective analysis.

SETTING

Using California hospital data, we examined the absolute and per admission changes in ED beds and inpatient beds in all hospitals from 2005 to 2014.

PARTICIPANTS

Our sample consisted of all patients inpatient and outpatient) from 501 hospital facilities over 10-year period.

OUTCOME MEASURES

We analysed linear trends in the total annual ED visits, ED beds, licensed and staffed inpatient hospital beds and bed types, ED beds per ED visit, and inpatient beds per admission (ED and non-ED).

RESULTS

Between 2005 and 2014, ED visits increased from 9.8 million to 13.2 million (an increase of 35.0%, p<0.001). ED beds also increased (by 29.8%, p<0.001), with an average annual increase of 195.4 beds. Despite this growth, ED beds per visit decreased by 3.9%, from 6.0 ED beds per 10 000 ED visits in 2005 to 5.8 beds in 2014 (p=0.01). While overall admission numbers declined by 4.9% (p=0.06), inpatient medical/surgical beds per visit grew by 11.3%, from 11.6 medical/surgical beds per 1000 admissions in 2005 to 12.9 beds in 2014 (p<0.001). However, there were reductions in psychiatric and chemical dependency beds per admission, by -15.3% (p<0.001) and -22.4% (p=0.05), respectively.

CONCLUSIONS

These trends suggest that, in its current state, inadequate supply of ED and specific inpatient beds cannot keep pace with growing patient demand for acute care. Analysis of ED and inpatient supply should capture dynamic variations in patient demand. Our novel 'beds pervisit' metric offers improvements over traditional supply measures.

A single-center cytogenetic study of 629 Chinese patients with de novo acute myeloid leukemia--evidence of major ethnic differences and a high prevalence of acute promyelocytic leukemia in Chinese patients

Cytogenetic information is important in the diagnosis, classification, and prognostication of acute myeloid leukemia (AML). Data obtained from multicenter treatment trials are well published. In this study, we contribute cytogenetic data from a large series of 629 Chinese patients with de novo AML that were karyotyped in a single laboratory. A higher prevalence of acute promyelocytic leukemia was observed when compared with non-Chinese series. The difference was most prominent in the younger age group. Abnormalities at chromosomal region 11q23 and inv(16) seemed uncommon. These ethnic differences may indicate underlying genetic susceptibility to AML development and/or environmental differences. More comprehensive data on AML in the elder population are needed to assess the role of cytogenetics in predicting prognosis and guiding treatment in this large subgroup of patients.

NMRES: an artificial intelligence expert system for quantification of cardiac metabolites from 31phosphorus nuclear magnetic resonance spectroscopy

The application of high-resolution 31Phosphorus Nuclear Magnetic Resonance (31P NMR) Spectroscopy in biology and medicine has provided new insights into biochemical processes and also a unique assessment of metabolites. However, accurate quantification of biological NMR spectra is frequently complicated by: (a) non-Lorentzian form of peak lineshapes, (b) contamination of peak signals by neighboring peaks, (c) presence of broad resonances, (d) low signal-to-noise ratios, and (e) poorly defined sloping baselines. Our objectives were to develop an expert system that captures and formalizes 31P NMR spectroscopists' expert knowledge, and to provide a reliable, efficient, and automated system for the interpretation of biological spectra. The NMR Expert System (NMRES) was written in the C and OPS5 programming languages and implemented on a Unix-based (Ultrix) mainframe system with XWindows bit-map graphics display. Expert knowledge was acquired from NMR spectroscopists and represented as production rules in the knowledge base. A heuristic weights method was employed to determine the confidence levels of potential peaks. Statistical and numerical methods were used to facilitate processing decisions. NMR spectra obtained from studies of ischemic neonatal and immature hearts were used to assess the performance of the expert system. The expert system performed signal extraction, noise treatment, resonance assignment, intracellular pH determination, and metabolite intensity quantitation in about 10 s per 4 KB (kilobyte) spectrum. The peak identification success rate was 98.2%. Peak areas and pH estimated by the expert system compared favorably with those determined by human experts. We conclude that the expert system has provided a framework for reliable and efficient quantification of complex biological 31P NMR spectra.

Monitoring of transcutaneous carbon dioxide tension

The authors studied the analytic and clinical characteristics of transcutaneous carbon dioxide monitoring in acutely ill NICU infants. Most infants were premature. The authors used both developmental and commercially available equipment. Monitoring was performed at temperature settings of 44 degrees C. With commercially available equipment, it was found that the least-squares linear regression relationship (r = 0.86, Sy.x = 12.9 torr, n = 100 patients) between transcutaneous PCO2 (y) and arterial PCO2 (x) was: y = -2.8 + (1.86)(x) (in torr). Calibration drift was unacceptably large in about 10% of the 500 monitoring sessions performed. Therefore, the authors conclude that transcutaneous PCO2 monitoring is most appropriate for following short-term trends in PCO2 over a period of one to three hours. Continuous transcutaneous PCO2 monitoring was most useful clinically after extubation when the objectives were careful observation for changes in PCO2 and prevention of reintubation.

Transcutaneous carbon dioxide for short-term monitoring of neonates

We studied transcutaneous pCO2 monitoring in 70 neonates, most of them premature with respiratory distress syndrome. Measurements were at 44 degrees C. Calibration drift was large in some instances. Least squares linear regression analyses of transcutaneous pCO2 (y) vs arterial pCO2 (x) in kilopascals showed, for all observations (n = 516), for one observation randomly selected from each patient (n = 70), and for the first observation from each patient (n = 70): y = -0.28 + 1.80x, y = 0.01 + 1.74x, and y = 0.73 + 1.63x, respectively. Regression lines for individual patients with 14 or more observations each were not coincident (F = 2.80, p less than 0.002). Transcutaneous pCO2 monitoring was most useful clinically as a means of following short-term trends in arterial pCO2 continuously during extubation and afterward when avoiding re-intubation. In view of the potential for error associated with drift, we recommend that intervals between calibrations be limited to about 3 h.

Transcutaneous carbon dioxide for short-term monitoring of neonates

We studied transcutaneous pCO2 monitoring in 70 neonates, most of them premature with respiratory distress syndrome. Measurements were at 44 degrees C. Calibration drift was large in some instances. Least squares linear regression analyses of transcutaneous pCO2 (y) vs arterial pCO2 (x) in kilopascals showed, for all observations (n = 516), for one observation randomly selected from each patient (n = 70), and for the first observation from each patient (n = 70): y = -0.28 + 1.80x, y = 0.01 + 1.74x, and y = 0.73 + 1.63x, respectively. Regression lines for individual patients with 14 or more observations each were not coincident (F = 2.80, p less than 0.002). Transcutaneous pCO2 monitoring was most useful clinically as a means of following short-term trends in arterial pCO2 continuously during extubation and afterward when avoiding re-intubation. In view of the potential for error associated with drift, we recommend that intervals between calibrations be limited to about 3 h.

Unpredictable fluctuations in transcutaneous pCO2 from capillary blood gas determinations

Transcutaneous pCO2 electrode response time was optimized by use of a new electrode filling solution composed of NaCl/NaHCO3 electrolyte buffer (100 and 20 mmol/L, respectively) in an equivolume mixture of glycerol and water. The 95% response time to a step change in pCO2 was 49.9 +/- 2.8 s (mean +/- SD) when there was no spacer between the membrane and glass of the electrode. Use of this filling solution during monitoring of severely ill premature infants with cardiopulmonary disease allowed identification of large, unpredictable transient changes in transcutaneous pCO2, and therefore presumably in arterial pCO2, that occurred during capillary blood gas sampling. The changes, which occurred in 19 of 20 samplings, ranged from -1.06 + 2.53 kPa (-8 to +19 Torr). The maximum relative change observed was +29%. These results indicate that the standard protocol for capillary blood collection induces significant transient fluctuations in blood gas tensions. We believe these fluctuations decrease the reliability of capillary pCO2 values for use in clinical management in patient populations similar to ours.