Brief Report

Increased rate of central venous catheterization procedures in community EDs^?

Seth W. Glickman MD ^a,c,?, Christopher Krubert MD ^b, Joel Koppenhaver BS ^b, Lawrence T. Glickman VMD, DrPH ^c, Kevin A. Schulman MD ^a, Charles B. Cairns MD ^a,c

^aDuke Clinical Research Institute, Duke University School of Medicine, PO Box 17969, Durham, NC 27715, USA

^bApolloMD Emergency Services, Atlanta, GA 30328, USA

^cDepartment of Emergency Medicine, University of North Carolina, Chapel Hill, NC 27599, USA

Received 4 August 2008; revised 13 October 2008; accepted 14 October 2008

Abstract

Objective: Central venous catheterization (CVC) is integral to the emergency department (ED) treatment of critically ill patients, such as those receiving early goal-directed therapy for severe sepsis. No previous studies have described the overall use of CVC in community EDs. The objective of this study was to estimate the overall frequency and temporal trends in CVC use in a sample of patients visiting community EDs.

Methods: This was a retrospective observational study of 2.97 million patient visits at 28 community EDs (range of annual visits, 10 837-110 136) from January 2004 to February 2008. Data were obtained from a community-based research consortium. Central venous catheterization procedures were aggregated at the hospital level for each study year. Trends in CVC use were evaluated using linear regression.

Results: Three thousand four hundred eighty-nine patient visits (0.12% of all ED patient visits) had a CVC procedure performed in the ED. The overall rate of CVC procedures per 1000 ED patient visits increased from 0.87 (95% confidence interval [CI_95%], 0.80-0.95) in 2004 to 1.62 (CI_95%, 1.38-1.91) procedures in 2008 (P value for trend = .003). There was wide variability in the frequency of CVC procedures performed among EDs, ranging from a low of 0.27 (CI_95%, 0.18-0.42) to a high of 7.58 (CI_95%, 6.27-9.17) procedures per 1000 ED visits. The CVC procedure rates were lower in the 8 rural EDs (0.99 CVCs per 1000 ED patient visits [CI_95%, 0.91-1.07] compared with the 20 urban EDs (1.22 CVCs [CI_95%, 1.18-1.27]; P b .001). An increasing rate of CVC procedures during the study period was observed in urban EDs (0.84-1.94 CVCs per 1000 ED patient visits; P value for trend = .005) but not in rural EDs (1.1-0.93; P value for trend = .41) during the study period.

Conclusion: The overall rate of CVC increased from 2004 to 2008. However, there was a wide variation among Eds, and the CVC rate was lower in rural compared with urban EDs. The increase in CVC use in urban EDs may reflect more intensive therapy in the management of ED patients with acute illness or injury. Future efforts are needed to optimize best practices for the use of CVC in community ED practices and to characterize factors responsible for urban rural differences in the rate of CVC procedures.

(C) 2010

^? Dr. Glickman’s role in this research was supported by the Douglas and Stephanie Kahn Charitable Gift Fund.

* Corresponding author.

E-mail address: [email protected] (S.W. Glickman).

0735-6757/$ – see front matter (C) 2010 doi:10.1016/j.ajem.2008.10.016

Introduction

Approximately 5 million central venous catheteriza- tion (CVC) procedures are performed annually in the United States [1]. Central venous catheterization is considered a core procedural competency of emergency physicians and is integral to the treatment of critically ill patients [2]. For example, CVC facilitates initial fluid resuscitation, delivery of vasoactive drugs, and assess- ment of hemodynamic status in emergency department (ED) patients with septic shock, a condition affecting more than 500 000 ED patients annually [3]. Each of these critical interventions are class I recommendations from the International Surviving sepsis campaign guidelines Committee for treatment of patients with severe sepsis [4].

Despite the important role CVC plays in the care of critically ill patients in the ED, no previous studies have described overall CVC use patterns in community ED practices. There are more than 5000 community EDs located in the United States, and greater than 90% of the 115 million ED patients seen annually are treated in a community practice setting [5]. The objective of this study was to determine the overall frequency of CVC from 2004 to 2008, and indications for CVC using a database created by a research consortium of 28 community EDs located primarily in southeastern states in the United States. In addition, interhospital variability was measured in rates of CVC procedures performed and temporal trends compared between urban and rural EDs.

Methods

This study was exempted from formal review by the institutional review board of XXXX. Data for this study were obtained from a central database created by the XXXX community-based research consortium of 28 EDs (range of annual visits, 10 837-110 136 per ED) that are part of the ApolloMD (Atlanta, Ga) physician group practice located primarily in the southeastern United States. Trained data coders at ApolloMD routinely abstract clinical and demographic data for all ED visits using standardized methods. At the time of this study, clinical data from a total of 28 ApolloMD hospitals were accessible for research purposes and were included in this analysis. Included in this study were all patients who had a central line placed in the ED (defined as Current Procedural Terminology code 36556) at any one of the 28 hospitals served by ApolloMD. The associated primary diagnoses (defined by the International Classification of Diseases, Ninth revision [ICD-9]) were identified for each patient visit. Data on CVC use were collected for the period January 2004 to February 2008. Data were analyzed independently at XXXX.

• 1. Statistical analysis

We conducted a retrospective observational study of CVC procedure use in community EDs from 2004 to 2008. The CVC procedure volume was aggregated yearly for each hospital. Procedure rates for CVC were calculated as the proportion of total ED patient visits associated with a CVC procedure. Overall trends in CVC use from 2004 to 2008 were evaluated for statistical significance using ordinary least squares linear regression in which the independent variable was year and the dependent variable was the proportion of patient visits that had a CVC procedure, weighted by the total number of patients treated in a given year.

Possible associations between hospital characteristics, such as ED volume and location (urban vs rural), and CVC rates were evaluated. Hospital characteristics and location were obtained from the database of the American Hospital Association. Hospitals were designated as rural or urban according to metropolitan statistical areas created by the United States Office of Management and Budget. Urban was defined as a geographical unit consisting of an urbanized area with a central city of at least 50 000 residents and a regional population of 100 000. Overall rates of CVC were compared between urban and Rural hospitals using the ?2 test. Overall trends in CVC from 2004 to 2008 by hospital type (rural and urban) were evaluated using weighted linear regression. Linear regression was also used to evaluate the possible association of ED volume with the rate of CVC use. The diagnoses of patients who received CVC were grouped into disease categories according to ICD-9 classi- fication. The number and percentage of patients who received CVC in each disease category were calculated. The most common diagnoses in each disease category were abstracted, and the CVC rate for each diagnosis was determined. The CVC rates for each disease category were

calculated for rural and urban hospitals separately.

All analyses were conducted with SAS statistical soft- ware, version 9.3.1 (SAS Institute Inc, Cary, NC); a P value of less than .05 was considered statistically significant.

Results

From January 2004 to February 2008, 2.97 million ED patient visits were made to the 28 study hospitals. Of the 28 hospitals, 27 (96.4%) were located in the southeast and 1 (3.6%) in the midwest. The mean hospital bed size was 258 (SD, 27), and the mean annual ED census was 36 873 patients (SD, 4384; range, 10 837-110 136 patients). Of the hospitals, 8 (29%) were located in nonmetropolitan, rural areas. These rural hospitals accounted for 649 229 patient visits (21.8% of all patient visits over the study period). The remaining 20 hospitals (71%) were located in urban areas and accounted for 2 324 994 patient visits (78.2% of all patient visits over the study period).

A total of 3489 patient visits (0.12% of all ED patient visits) underwent a CVC procedure in the ED. The overall rate of CVC per 1000 ED patient visits increased from 0.87 (95% confidence interval [CI_95%], 0.80-0.95) in 2004 to 1.62 (CI_95%, 1.38-1.91) procedures in 2004 (R² = 0.96; P value for trend = .003) (Table 1). There was wide variability in CVC rates between hospitals, ranging from 0.27 (CI_95%, 0.18-0.42) to 7.58 (CI_95%, 6.27-9.17) CVCs per 1000 ED

Table 1 Rate and CI_95% of CVC procedures performed per 1000 ED patient visits in 28 community EDs from January 2004 to February 2008

visits (Fig. 1). Overall procedure rates were lower in the 8 rural EDs (0.99 CVCs per 1000 ED patient visits [CI_95%, 0.91-1.07]) compared with the 20 urban EDs (1.22 CVCs per 1000 ED patient visits [CI_95%, 1.18-1.27]; P value for difference b .001) (Table 1). Rates in CVC use increased significantly over the study period at urban EDs (0.84-1.94 CVCs per 1000 ED patient visits, R² = 0.95; P value for trend = .005) but not in rural EDs (1.1-0.93 CVCs per 1000 ED patient visits, R² = 0.05; P value for trend = .41) (Table 1). The ED volume was not significantly associated with the rate of CVC use (R² = 0.003; P = .79).

The most common Diagnostic categories of patients undergoing CVC procedures were circulatory (25.4% of all CVCs), respiratory (13.2%), digestive (13.1%), and infec- tious (12.4%) (Table 2). The most common individual diagnosis among patients undergoing CVC were cardiac

arrest (13.4%), acute respiratory failure (13.2%), sepsis and septic shock (7.0%), unspecified shock and hypovolemia (6.1%), gastrointestinal bleeding (5.7%), acute coronary syndrome (5.6%), and traumatic injuries and burns (5.6%) (Table 1). Among patients who underwent CVC in the ED, the proportion of patients in each of the disease diagnostic categories was similar in rural and urban EDs.

Discussion

Central venous catheterization is considered a core competency of emergency physicians and a critical intervention in the management of a wide variety of ED patients with severe illness and injury [2]. Despite the large number of patients treated in community ED practices, few published information is available on the frequency of use of CVC in this setting. In the present study using a community-based research consortium of 28 hospitals with a total of 2.97 million ED patient visits, overall rates of CVC increased significantly from 2004 to 2008, but there was wide variability between hospitals. Hospitals located in rural areas had significantly lower volume-adjusted CVC rates compared with urban hospitals, and these rates

Fig. 1 Interhospital variability in the rate of CVC procedures performed. Figure 1 shows the rate of CVC procedures per 1000 ED visits at each of the 28 study hospitals from January 2004 to February 2008.

Table 2 Final diagnoses for 3489 patients visits with CVC placement in 28 EDs

CVC, n (%)

Total Rural Urban

did not increase significantly over time as was seen in urban areas.

The reason for the observed increase over time in the rate of CVC procedures in urban hospitals is likely multifactorial. One possible explanation is that more intensive therapy is being used in the management of patients with acute illness and injury in EDs. For example, more aggressive therapy is increasingly being used to manage patients with sepsis and septic shock in the ED. Several studies have demonstrated the benefit and cost-effectiveness of early goal-directed therapy for sepsis, and professional society guidelines have been disseminated to emergency physicians [6,7]. Because advanced training in emergency medicine includes compe-

tency in CVC, a possible explanation for the increase in the rate of CVC observed in this study may be a greater number of emergency medicine-trained and board-certified/eligible physicians entering emergency practices, particularly in urban areas.

Studies on the quality of clinical care provided in community EDs are needed because a large proportion of ED patient visits occur at more than 4600 EDs in the United States [5]. Overall CVC use in this study was significantly lower in EDs located in rural areas. Although the lower rate of CVC use in rural EDs could possibly be explained by differences in patient acuity and case-mix between urban and rural EDs, several findings from this study were not

supportive. First, in contrast to urban EDs, there was no increase in CVC rates in rural EDs over the study period. Second, the distribution of diagnoses for patients who received CVCs in rural EDs was similar to those in urban EDs. An alternative explanation for lower CVC rates in rural areas is that there are fewer emergency medicine-trained, board-certified/eligible physicians in rural compared with urban areas. In the urban EDs in this study, approximately 95% of the emergency medicine physicians were residency- trained and board-certified/eligible in emergency medicine vs approximately 65% of the physicians in the rural EDs. A recent study found that throughout the United States, non- emergency-trained physicians provided greater than half of all ED care in rural areas [8]. In addition, The American College of Physicians national workforce study conducted in 1999 found that 38% of ED physicians were not emergency medicine residency-trained or board-certified, including 67% of those in rural areas [9].

A better understanding of CVC use in community practice is important from a policy perspective. Beginning in October 2008, the Centers for Medicaid and Medicare Services will no longer provide additional reimbursement to hospitals for health care-associated infections. Therefore, infections acquired via any procedure performed in the ED, including CVC, will directly impact Hospital reimbursement. In addition, the use of a CVC insertion protocol that includes a maximum sterile barrier technique is 1 of 11 Centers for Medicaid and Medicare Services Physician Quality Report- ing Initiative performance measures for 2008. Given the large volume of CVCs placed in the ED, appropriate training and procedural capability of emergency providers who place CVCs in the community setting is necessary. In addition, ultrasonographic guidance during CVC has been associated with a lower complication rates and is a Joint Commission on the Accreditation of Health care Organizations adopted standard for CVC placement in the ED. Yet, a recent study found that greater that two thirds of community EDs report no access to ultrasonography by emergency physicians [10]. Greater availability of ultrasonography in community ED practice will improve patient safety during CVC.

There are several limitations to this study. First, ICD-9 classification codes may not capture important clinical information. For example, a primary diagnosis of acute respiratory failure may reflect a variety of causes including primary respiratory disease, traumatic injury, infection, and others. Second, the overall rates of CVC observed in this study may underestimate actual rates because instances when a physician performed CVC but did not document the procedure in the medical record would not be captured by data abstractors. Third, increased improvements in docu- mentation and coding of patient records could theoretically account for the increased rates observed in CVC use over time. However, there were no apparent systematic changes

made in CVC clinical protocols, physician documentation, or coding procedures at any of the study EDs during the period of the study. These factors would also not explain the differential increase in CVC procedures between urban and rural hospitals in this study. No significant changes occurred in physician reimbursement for CVC, and no major CVC Efficacy studies were published during the study period. Finally, although this study includes 28 EDs located primarily in the southeastern United States, future research is needed to evaluate urban-rural differences in procedures rates among a larger sample of community EDs across a larger Geographical area.

In conclusion, we found that overall rates for CVC procedures significantly increased from 2004 to 2008 in a sample of community EDs, particularly in urban areas. This change may reflect adoption of more intensive therapy in the ED for management of patients with acute and severe time sensitive illnesses and injury. Studies are needed to determine optimum CVC protocols in community ED practice including ultrasound-guided insertion and early goal-directed therapy for management of septic shock and hypothermia in Postcardiac arrest patients. Additional research is also needed to explain potential disparities in the quality of emergency and acute care in urban vs rural EDs.

References

1. Taylor RW, Plagiri AV. Central venous catheterization. Crit Care Med 2007;35(5):1390-6.
2. Society for Academic Emergency Medicine and Council of Emer- gency Medicine Residency Directors. Model Curriculum and Guide- lines for Curriculum Development for Emergency Medicine Residency Training. Accessed at http://www.saem.org/saemdnn/Education/tabid/ 55/Default.aspx on June 8, 2008.
3. Wang HE, Shapiro NI, Angus DC, Yealy DM. National estimates of severe sepsis in United States emergency departments. Crit Care Med 2007;35(8):1928-36.
4. Dellinger RP, Levy MN, Carlet JM, et al. Surviving Sepsis Campaign: International guidelines for management of severe sepsis and septic shock: 2008. Intensive Care Med 2008;34(1):17-60.
5. Kellermann AL. Crisis in the emergency department. N Eng J Med 2004;355:1300-3.
6. Talan DA, Moran GJ, Abrahamian FM. Severe sepsis and septic shock in the emergency department. Infect Dis Clin North Am 2008;22(1): 1-31.
7. Talmor D, Greenberg D, Howell MD, et al. The costs and cost- effectiveness of an integrated sepsis treatment protocol. Crit Care Med 2008;36(4):1168-74.
8. Peterson LE, Dodoo M, Bennett KJ, et al. Nonemergency medicine- trained physician coverage in rural emergency departments. J Rural Health 2008;24(2):183-8.
9. Moorhead J, Gallery MK, Hirshkorn C, et al. A study of the workforce in emergency medicine. Ann Emerg Med 2002;40(1):3-15.
10. Moore CL, Molina AA, Lin H. Ultrasonography in community emergency departments in the United States: access to ultrasonography performed by consultants and status of emergency physician- performed ultrasonography. Ann Emerg Med 2006;47(2):147-53.