Original Contributions
Factors Associated with Hospitalization among Emergency Department Patients Referred for Quick Investigation of Iron-Deficiency Anemia

https://doi.org/10.1016/j.jemermed.2015.08.010Get rights and content

Abstract

Background

Although patients with anemia are frequently seen in emergency departments (EDs), studies on patients presenting there with symptomatic chronic anemia—usually iron-deficiency anemia (IDA) caused by occult gastrointestinal bleeding—are lacking. Awareness of predictors of hospitalization could direct the ED triage to the appropriate diagnostic setting.

Objective

Based on initial observations that some patients with IDA were hospitalized after ED referral and initial evaluation at a quick diagnosis unit (QDU), a new cost-effective alternative to hospitalization for diagnostic workup, this study aimed to determine the patient factors associated with hospitalization after the first QDU visit.

Methods

An 8-year prospective cohort study of patients with IDA referred from the ED to the QDU of a third-level university hospital was conducted. Patients with a baseline hemoglobin level of <9 g/dL in the ED, proven iron deficiency, and no overt bleeding were included. The primary outcome was hospitalization after the initial QDU assessment.

Results

Two hundred eighty-four (7.7%) of 3692 patients were hospitalized. Inter-rater agreement of appropriateness of admissions was 90.6% (κ = 0.82). Overall, 90% of study patients presented to the ED with symptomatic anemia, and 87% were transfused there. On multivariate analysis, age ≥65 years, living alone, a post-transfusion hemoglobin level of <9 g/dL, higher age-adjusted overall comorbidity, heart failure, and poor physical health-related quality of life at first QDU visit independently predicted hospitalization.

Conclusion

While these predictors do not necessarily reflect the need for hospitalization, they are easily evaluated during the initial ED visit and can guide the triage of similar IDA patients to the suitable setting for timely investigation.

Introduction

Patients with anemia are frequently seen in emergency departments (EDs), and emergency physicians may play an essential role in their evaluation and initial management (1). However, data on the frequency of anemia in the general ED population are limited (2). Notably, apart from reports on the occurrence and management of anemia in pediatric, obstetric, and gynecologic (i.e., abnormal uterine bleeding) patients, ED patients with sickle cell disease, and ED management, including red blood cell (RBC) transfusion strategies in patients with acute gastrointestinal (GI) hemorrhage or in trauma and critical care, no studies on adult subjects presenting to EDs with symptomatic chronic anemia—most commonly corresponding to microcytic, iron-deficiency anemia (IDA) caused by occult GI bleeding—have been reported 1, 2, 3.

In Spain, a country with a publicly funded health system, physicians at primary care centers (PCCs) commonly refer patients with moderate to severe IDA (hemoglobin <10 g/dL) to EDs to gain quicker access to GI investigations via emergency admission 4, 5. While the standard operating procedure has been to hospitalize these patients, even the young ones, for speeding diagnostic workup, the recent creation of hospital-based outpatient quick diagnosis units (QDUs) has represented a shift in PCC referrals and ED admission practices (4). In patients with different potentially serious conditions, frequently associated with cancer, QDUs have proven a solution to diagnostic delays and an efficient, cost-saving alternative to hospitalization for diagnostic workup 4, 5, 6, 7, 8. While referral criteria for anemia vary among QDUs, our unit assesses anemic patients, mostly with new, uninvestigated IDA, with an index hemoglobin value measured at the time of presentation to the ED (“baseline hemoglobin”) of <9 g/dL and no acute GI bleeding. Although the optimal hemoglobin threshold warranting quick investigation of IDA has not been elucidated, the rationale for using this cutoff is evidence showing that risk of GI (mainly colorectal) malignancy in IDA patients is significantly increased when the hemoglobin level is <9 g/dL 9, 10, 11, 12. Initial observations showed, however, that some patients with IDA presenting to the ED with baseline hemoglobin <9 g/dL who were referred to QDU were unpredictably hospitalized shortly after starting the QDU assessment.

The delineation of risk factors has permitted us to systematize the care, including the appropriate management setting (inpatient vs. outpatient setting) of patients presenting with acute overt GI hemorrhage and secondary anemia, and improved decision-making at the early triage level has led to improved outcomes, resource use, and overall management (13). In contrast, factors associated with hospitalization of patients presenting to EDs with significant chronic IDA are unknown, meaning that early care and triage to the proper level of care for evaluating these patients for diagnostic purposes relies primarily on the judgment of referring physicians or on institution practices or guidelines.

To elucidate this evidence gap, the goal of this study was to determine, among ED adults referred to the QDU of a third-level university hospital for investigation of IDA, the patient factors associated with hospitalization after the initial QDU visit.

Section snippets

Study Design

This was a prospective cohort study of patients with IDA who were consecutively referred from the ED to the QDU between September 2005 and October 2013. The primary outcome was hospitalization after initial QDU consultation. The research ethics committee of the hospital approved the study, and informed consent was obtained from all patients.

Study Setting and Population

The QDU is located in an 870-bed third-level university hospital in Barcelona (Spain) with a reference population of almost 550,000. The main requirements

Results

Among 3814 initially eligible patients referred to the QDU who had proven IDA, 53 had patent blood losses from non-GI sources (e.g., profound menstruation in 33 premenopausal and metrorrhagia in 15 postmenopausal women), 38 were lost to follow-up, 17 refused to participate, and 14 died before outpatient investigations concluded. Analysis was restricted to 3692 patients, with a mean age of 65.2 (17.4) years, and 1911 (52%) patients were women. Some enrolled patients (2 hospitalized and 114

Discussion

This prospective study is, as far as we are aware, the first to examine the patient factors associated with admission among adults referred from the ED for prompt outpatient investigation of IDA. In our study, 7.7% of subjects with confirmed IDA with baseline hemoglobin levels in the ED of <9 g/dL were hospitalized after onward referral to and initial assessment at an outpatient diagnostic clinic that is expressly devised to rapidly investigate patients with potentially serious diseases.

Conclusions

Various patient factors were independently associated with admission among ED patients who were referred for rapid outpatient investigation of IDA. These predictive factors are easily available on the initial ED evaluation and can guide the triage to the most appropriate and safe setting for timely investigation. Although these independent predictors do not necessarily portray the need for admission, their knowledge may be useful for emergency physicians who encounter similar patients with IDA

References (33)

  • T.G. Janz et al.

    Anemia in the emergency department: evaluation and treatment

    Emerg Med Pract

    (2013)
  • J.M. Pericás et al.

    Should alternatives to conventional hospitalisation be promoted in an era of financial constraint?

    Eur J Clin Invest

    (2013)
  • X. Bosch et al.

    Quick diagnosis units: a potentially useful alternative to conventional hospitalization

    Med J Aust

    (2009)
  • X. Bosch et al.

    Evaluation of unexplained peripheral lymphadenopathy and suspected malignancy using a distinct quick diagnostic delivery model: prospective study of 372 patients

    Medicine (Baltimore)

    (2014)
  • W. Hamilton et al.

    The importance of anaemia in diagnosing colorectal cancer: a case-control study using electronic primary care records

    Br J Cancer

    (2008)
  • M.W. James et al.

    Risk factors for gastrointestinal malignancy in patients with iron-deficiency anaemia

    Eur J Gastroenterol Hepatol

    (2005)
  • Cited by (5)

    • A comprehensive evaluation of the gastrointestinal tract in iron-deficiency anemia with predefined hemoglobin below 9 mg/dL: A prospective cohort study

      2017, Digestive and Liver Disease
      Citation Excerpt :

      Providing that physical performance is well enough to travel to hospital for tests and visits, patients with unexplored anemia (almost always corresponding to microcytic anemia with suspected iron deficiency) with hemoglobin values <9 g/dL are currently referred from the emergency departments (EDs) and primary care centers (PCCs) to our hospital for prompt evaluation by QDU physicians [22,23]. Previous studies have shown that in comparison with most IDA patients uneventfully assessed at QDUs following ED referral, other patients are more likely to be hospitalized soon after the first QDU visit, before conducting GI procedures, due to clinically symptomatic anemia, high fatigue, and comorbidities (mainly cardiovascular disease) associated with significantly lower hemoglobin values [24]. To determine whether the underlying causes of IDA could explain these different outcomes, we undertook a prospective study to investigate GI disorders accounting for IDA in a large cohort of patients referred to QDUs with baseline hemoglobin values <9 g/dL, iron deficiency, and no extraintestinal causes of blood loss or iron deficiency.

    Reprints are not available from the authors.

    View full text