Epidemiology and Healthcare Costs of Incident Clostridium difficile Infections Identified in the Outpatient Healthcare Setting

Jennifer L. Kuntz, PhD; Eric S. Johnson, PhD; Marsha A. Raebel, PharmD; Amanda F. Petrik, MS; Xiuhai Yang, MS; Micah L. Thorp, DO, MPH; Steven J. Spindel, MD; Nancy Neil, PhD; David H. Smith, PhD


Infect Control Hosp Epidemiol. 2012;33(10):1031-1038. 

In This Article


We conducted this study to identify and describe patients with CDI diagnosed in the outpatient setting, as well as to characterize CDI-related healthcare utilization and costs among these patients. We also compared baseline characteristics of persons with outpatient- and inpatient-identified CDI to help clinicians and researchers better understand how patients with CDI infections identified in the 2 settings may differ. We found that slightly more than one-half of incident CDIs were identified in the outpatient setting. These patients tended to be younger with less comorbidity than patients with CDI identified in the inpatient setting. Collectively, our results emphasize that CDI is being identified and treated among younger, healthier ambulatory populations.

Our study categorized patients by the setting in which infection was first identified rather than the setting where C. difficile was likely acquired (ie, community associated, healthcare facility associated). We acknowledge that widely used surveillance approaches that categorize infections by location in which a patient may have been exposed to C. difficile are important in identifying outbreaks and designing prevention efforts. However, the setting in which CDI is identified represents the first opportunity for clinicians to diagnose and intervene on infections. As such, our study is important for understanding characteristics of the growing number of patients presenting with CDI in the outpatient setting. In addition, our results suggest that a large number of CDIs are not captured by current efforts focused on CDI among hospitalized patients. This knowledge can inform efforts to improve timely clinical recognition and treatment of infection among outpatients, thus potentially preventing prolonged illness, adverse outcomes, and additional healthcare utilization.

We found few apparent differences between patients with CDI in either setting, with the exception that patients with outpatient-identified CDI were younger and had lower comorbidity load. These observations could be a reflection of the differences in the demographic and clinical characteristics of ambulatory and hospitalized patients rather than a difference in epidemiology between the 2 settings. We also utilized a logistic regression model to compare persons with infections identified in the 2 settings. Almost all of the variables in our model were associated with identification of CDI among hospitalized patients rather than among outpatients. However, we did find that collectively, the 21 patient characteristics in the model adequately discriminated the setting in which CDI was identified (c-statistic, 0.76). This suggests that patients who seek outpatient care for CDI differ from patients with CDI identified in the inpatient setting, although it may be difficult for providers to anticipate the setting in which a patient will present with a CDI because the individual predictors are so subtle. Finally, because our study did not include a population without CDI, these results should not be used to infer any statistical differences between populations with CDI and those without.

A substantial proportion of CDI patients in our population were exposed to antimicrobials before identification of their infection. Still, 14% of patients with outpatient-identified CDI had not received an antimicrobial in the 180 days before CDI. In addition, although prior research has suggested that the majority of CDI in the outpatient setting may be attributed to inpatient exposures,[18] only 27% of our patients with outpatient-identified CDI had a history of hospitalization in the previous 30 days, the period of time used by the Centers for Disease Control and Prevention to define community-onset, healthcare facility–associated CDI. Taken together, our observations reinforce the fact that CDI is occurring among ambulatory patients who may be considered at low risk. Furthermore, these results suggest that a substantial proportion of these infections not only were identified in the outpatient setting but also were acquired there.

In our population, the impact of CDI on healthcare utilization and cost was most notable in the setting in which the patient's infection had been identified. For example, patients with outpatient-identified CDI were relatively more likely to seek additional outpatient care, while patients with inpatient-identified CDI were more likely to experience additional hospitalization. However, we found that patients with inpatient-identified CDI received a greater number of treatment dispensings from outpatient pharmacies even after their initial hospitalization, suggesting that disease was unresolved during hospitalization and illness was prolonged after hospital discharge.

Our study provides insight into the patterns of care for patients with CDI, although we must clarify that the cost estimates reported here should be viewed as ceiling estimates because they include visits with joint production of care. For example, the entire cost of care for a patient who experiences a CDI during a hospitalization for heart failure is reported, even though the patient will have also received heart failure care unrelated to CDI. Other investigators[19] have shown that about 30% of total hospital care costs for patients with concomitant CDI are attributable to the infection; thus, we might estimate that about one-third of the cost reported by our study is directly attributable to C. difficile. On the other hand, any infection control procedures related to CDI (eg, isolation) that may have been implemented in the hospital setting are not represented in our inpatient cost estimates.

Our study has a number of limitations. First, because of data limitations, we collected only information about outpatient prescription dispensings; as a result, we underestimate medication exposures among hospitalized patients. In addition, because gastric acid suppressants are available over-the-counter, we could not measure use among patients who did not acquire these medications through prescription. Second, we identified patients with CDI in the inpatient setting through ICD-9 codes only. Although it would have been optimal to obtain results for C. difficile toxin testing from this population, the ICD-9 code for CDI has reasonable sensitivity and specificity for detecting cases in inpatient settings.[13,14] It is possible that CDIs identified through ICD-9 codes during a hospitalization may actually represent patients with a history of CDI. However, we believe that we minimized this possibility by requiring that patients have no evidence of CDI in the previous 180 days and by using only initial infections in our analysis. Moreover, we could not determine the date on which symptoms first occurred or CDI was diagnosed during a hospitalization; thus, data collection for these cases is based on the admission date of the hospitalization during which CDI was diagnosed. In contrast, we used both ICD-9 codes and toxin test results to identify CDI in the outpatient setting and required patients with a positive toxin test result to also have a C. difficile diagnosis or evidence of CDI treatment. Thus, we likely identified the majority, if not all, of the cases occurring in the outpatient setting. Finally, although we were able to measure potential risk factors (eg, patterns of antimicrobial use) among the patients in this study, we do not know whether they differ from background rates of exposures and patient characteristics of the general outpatient and inpatient populations.

Our study documented that slightly more than one-half of all CDIs occurring in our population were identified in the outpatient setting; thus, we conclude that CDI demands greater attention in the outpatient setting. Our results suggest that obvious risk factors for CDI in the outpatient setting may be lacking; nevertheless, clinicians should obtain appropriate diagnostic testing on outpatients with potential CDI.