Rat-Bite Fever in the United States: An Analysis Using Multiple National Data Sources, 2001–2015

Abstract

Background

Rat-bite fever is a rare disease associated with rat bites or direct/indirect rodent contact.

Methods

We examined rat-bite fever and rat-bite injury diagnoses in the United States during 2001–2015. We analyzed national, state, and Indian Health Service healthcare encounter datasets for rat-bite fever and rat-bite injury diagnoses. We calculated average-annual encounter rates per 1 000 000 persons.

Results

Nationally, the rat-bite fever Emergency Department visit rate was 0.33 (95% confidence interval [CI], 0.19–0.47) and the hospitalization rate was 0.20 (95% CI, 0.17–0.24). The rat-bite injury Emergency Department visit rate was 10.51 (95% CI, 10.13–10.88) and the hospitalization rate was 0.27 (95% CI, 0.23–0.30). The Indian Health Service Emergency Department/outpatient visit rate was 3.00 for rat-bite fever and 18.89 for rat-bite injury. The majority of rat-bite fever encounters were among individuals 0–19 years of age.

Conclusions

Our results support the literature that rat-bite fever is rare and affects children and young adults. Targeted education could benefit specific risk groups.

Most rat-bite fever healthcare encounters were among children, young adults, and American Indian/Alaska Native populations. Although rare, rat-bite fever occurs at a higher rate than is demonstrated in existing literature. Targeted education could benefit specific risk groups.

Rat-bite fever (RBF) is a bacterial zoonotic disease caused primarily by Streptobacillus moniliformis [1–4]. Exposure to rodent and nonrodent species have resulted in human infections; however, specific mechanisms of bacterial transmission are not well understood [5, 6]. Classic case reports cite bites or scratches from rats or other rodents; however, recent literature describes human infection after handling animals without a distinct wound or through fomite transmission. Cases are predominantly documented among children and are increasingly associated with pet rodent ownership [7–9].

The incubation period for RBF ranges from 3 to 21 days; however, most patients develop symptoms 3 to 10 days after exposure [10, 11]. The disease can produce a broad range of symptoms. Patients may initially present with headache, fever, chills, vomiting, and severe myalgia. Disease progression often results in migratory polyarthritis and may include rash; complications include focal abscesses, hepatitis, nephritis, pneumonia, meningitis, endocarditis, and death [12]. Nonspecific and varied clinical presentation make prompt RBF diagnosis difficult, particularly without a comprehensive exposure history [13]. Rat-bite fever is diagnosed by culture isolation of S moniliformis; however, in the absence of a positive culture, Gram stain identification in appropriate specimens can support a preliminary diagnosis. Streptobacillus moniliformis is generally susceptible to several antibiotics. Treatment commonly includes penicillin, with doxycycline or streptomycin as alternative choices [14–16].

Rat-bite fever incidence in the United States is unknown. The disease is not reportable in any state or territory, nor is it a nationally notifiable condition. Consequently, epidemiological data are limited to published case reports documenting the clinical course of patients. To gain national perspective on this neglected disease, we estimated the number of RBF hospitalizations, emergency department (ED) visits, and outpatient clinic visits across 15 years using retrospective hospital data. We analyzed rat-bite injury (RBI) hospitalizations, ED, and outpatient clinic visits to assess the magnitude of RBF within the context of RBIs.

METHODS

Data Sources

We analyzed 5 datasets for RBF and RBI to compare hospitalizations, ED, and outpatient clinic visits (together referred to as “encounters”) across datasets with different sampling strategies, units of analysis, age of the sample/population, and years of data availability (Table 1). We obtained 4 datasets from the Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality, including the following: (1) National (Nationwide) Inpatient Sample (NIS); (2) Kids’ Inpatient Database (KID); (3) the Nationwide Emergency Department Sample (NEDS); and (4) the State Inpatient Database (SID) [17]. In addition, we analyzed Indian Health Service (IHS) Direct and Contract Health Service Inpatient and Outpatient Datasets, from the National Patient Information Reporting System (NPIRS) [18].

Table 1.

Data Sources for Rat-Bite Fever and Rat-Bite Injury Encounters^a, United States,

Data Source	Dataset	Population or Sample	Unit(s) of Analysis	Ages	Years
Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality	National (Nationwide) Inpatient Sample (NIS)	Sample	Hospitalization	All	2001–2015
	Kids’ Inpatient Database (KID)	Sample	Hospitalization	≤20 years	2003, 2006, 2009, 2012
	Nationwide Emergency Department (ED) Sample (NEDS)	Sample	ED visit	All	2006–2015
	State Inpatient Database (SID)b	Population	Hospitalization	All	2001–2015
National Patient Information Reporting System, Indian Health Service (IHS)	Direct and Contract Health Service Inpatient and Outpatient Datasets	Population	Hospitalization; ED or outpatient visit	All	2001–2015

Abbreviations: ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification.

^aAn encounter is defined as any hospitalization, ED, or outpatient visit.

^bA total of 31 states contributed data to the SID during 2001–2015, with the number of participating states varying each year. Annually, an average of 24.5 states contributed data during the study period, ranging from 8 to 25. The US Census Bureau defines 4 regions in the United States (East, South, Midwest, and West); between 50% and 69% of states within each region participated during the study period.

Healthcare Cost and Utilization Project yields the largest all-payer hospital care database in the United States, providing nationally representative estimates of hospitalizations and hospital-based ED visits. Within HCUP, NIS provides estimates based on all discharges from the following: (1) a 20% stratified sample of community hospitals, with approximately 1000 facilities (2001–2011) [19]; or (2) a 20% stratified sample of discharges from community hospitals, excluding rehabilitation and long-term acute care hospitals (2012–2015) [20]. The KID data are generated every 3 years through a sample of pediatric discharges from community hospitals, excluding rehabilitation hospitals [21]. Nationwide Emergency Department Sample captures ED visits using a 20% stratified sample of hospital-based EDs, with information from approximately 30 million unweighted ED visits at 1000 hospitals [22]. Finally, SID contains 100% of inpatient records from contributing states (Supplementary Figure 1) [23]. Both NIS and KID were developed based on samples drawn from SID, with a sampling frame stratified by the number of beds, teaching status, ownership, rural/urban location, and region of the facility [19-21].

Indian Health Service NPIRS Direct and Contract Health Service Inpatient and Outpatient datasets are aggregated from IHS federal or tribally operated hospitals and clinics as well as community hospitals and clinics contracted by tribes or IHS to provide healthcare services to eligible American Indian and Alaska Native (AI/AN) patients [24]. We aggregated IHS data into 7 geographic regions: East, Northern Plains East, Northern Plains West, Southern Plains, Southwest, West, and Alaska (Supplementary Figure 2).

Numbers and rates across HCUP data sources should be examined independently and not aggregated. There is probable overlap between hospitalizations in NIS, KID, and SID, and a hospitalization in one of these data sources could have originated as an ED visit in NEDS. Although minimal, there may also be overlap between IHS and HCUP data.

Definitions

We defined a RBF encounter as an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) [25] code “026” or ICD-10-CM [26] code “A25” in any one of up to 25 diagnoses listed on the patients’ encounter record. We defined an RBI encounter as ICD-9-CM code “E906.1” in any one of up to 15 diagnoses listed on the patients’ encounter record in pre-2003 HCUP datasets. After 2002 in HCUP and for all IHS data, RBI diagnosis was defined as ICD-9-CM code “E906.1” or ICD-10-CM code “W53.11” in any one of up to 12 injury codes listed on the encounter record. Together, we refer to these diagnoses as rat-bite-related encounters (Table 1).

Statistical Analysis

We examined RBF and RBI encounters across age groups (0–19, 20–39, 40–59, and ≥60 years), sex, region, urbanicity (classified by county of residence), race, primary payer, length of stay (LOS) in the facility, and ED patient outcome, based on the information available in each dataset.

Due to the sampling design of NIS, KID, and NEDS, we used SUDAAN software to calculate and report weighted hospitalization/ED estimates, percentages, and their standard errors (SEs) [27]. We only performed analyses when the number of unweighted hospitalizations/ED visits per strata was >10 and the relative SE of the estimates was ≤0.30 (NIS, KID, NEDS), the number of hospitalizations per strata was >10 (SID), or the number of encounters per strata was ≥5 (IHS).

We used bridged-race population estimates from the National Center for Health Statistics as the population denominator for HCUP data to calculate rates [28]. For NIS, KID, and NEDS, we expressed annual and average-annual hospitalization/ED visit rates as the number of weighted hospitalizations/ED visits per 1 000 000 persons with 95% confidence intervals (CIs). For SID, we used census population estimates for each year that a state participated to calculate person-year denominators and calculated hospitalization rates per 1 000 000 person-years.

Within IHS NPIRS, we expressed average-annual encounter rates as the number of encounters per 1 000 000 persons. We used the IHS user population as the denominator for rates, which included individuals who received IHS-funded healthcare services at least once during the previous 3 years [24, 29]. Emergency department rates exclude the West region that do not have IHS or tribally operated EDs and contract health service data are incomplete.

We calculated SEs for the number of encounters and percentages and CIs of rates for samples. We calculated rate ratios (RRs) using Poisson regression, but only for complete datasets (SID and IHS) due to unreliability in the sample estimates. We calculated median LOS and associated SEs (when applicable) for RBF and RBI overall and by age group, sex, and race/ethnicity for NIS and SID, region (NIS), and state (SID). Finally, we calculated median charges per hospitalization for SID.

The use of HCUP data prohibits identifying individuals. This study was exempt from CDC Institutional Review Board approval. We conducted all analyses using SAS statistical software (version 9.4; SAS Institute, Inc., Cary, NC) and SUDAAN where applicable. Statistical significance was set at P < .05.

RESULTS

National (Nationwide) Inpatient Sample

In NIS, there were 929 (SE, 73) RBF hospitalizations in the United States from 2001 to 2015, with an average-annual rate of 0.20 per 1 000 000 persons (95% CI, 0.17–0.24) (Table 2). The rate of RBF hospitalizations was highest among those 0–19 years of age. A majority occurred among males, persons of white race, urban residents (reported 2007–2015), and private insurance holders, including those under Health Maintenance Organizations (HMOs).

Table 2.

Weighted Number, Percentage, and Average-Annual Rates of Rat-Bite Fever and Rat-Bite Injury Hospitalizations by Demographic Attribute and Payment Type, NIS, United States, 2001–2015^a

	RBF			RBI
Characteristic	Weighted No. ± SE	Percent ± SE	Rate (95% CI)b	Weighted No. ± SE	Percent ± SE	Rate (95% CI)b
Overall	929 ± 73	100	0.20 (0.17–0.24)	1216 ± 81	100	0.27 (0.23–0.30)
Age Group (Years)
0–19	480 ± 54	52.2 ± 3.8	0.39 (0.30–0.48)	255 ± 37	21.0 ± 2.7	0.21 (0.15–0.27)
20–39	208 ± 32	22.7 ± 3.1	0.17 (0.12–0.22)	338 ± 40	27.8 ± 2.8	0.27 (0.21–0.34)
40–59	143 ± 26	15.6 ± 2.6	0.11 (0.07–0.16)	367 ± 43	30.2 ± 3.0	0.29 (0.23–0.36)
60+	88 ± 22	9.6 ± 2.3	0.11 (0.05–0.16)	255 ± 37	21.0 ± 2.7	0.31 (0.22–0.40)
Sex
Male	495 ± 53	54.2 ± 3.7	0.22 (0.18–0.27)	515 ± 51	42.4 ± 3.1	0.23 (0.19–0.28)
Female	419 ± 46	45.9 ± 3.7	0.18 (0.14–0.22)	701 ± 60	57.6 ± 3.1	0.30 (0.25–0.35)
Race/Ethnicity
White	596 ± 57	82.9 ± 3.1	0.20 (0.16–0.24)	635 ± 58	64.8 ± 3.6	0.21 (0.17–0.25)
Black	-	-	-	151 ± 28	15.4 ± 2.6	0.26 (0.17–0.36)
Hispanic	53 ± 16	7.3 ± 2.1	-	126 ± 26	12.9 ± 2.4	0.18 (0.11–0.25)
Otherc	-	-	-	68 ± 18	6.9 ± 1.8	NR
Region
Northeast	172 ± 29	18.5 ± 2.9	0.21 (0.14–0.28)	299 ± 42	24.6 ± 3.0	0.36 (0.26–0.46)
Midwest	253 ± 38	27.3 ± 3.5	0.25 (0.18–0.33)	181 ± 31	14.9 ± 2.3	0.18 (0.12–0.24)
South	234 ± 37	25.3 ± 3.5	0.14 (0.10–0.18)	480 ± 50	39.5 ± 3.2	0.29 (0.23–0.35)
West	269 ± 41	29.0 ± 3.6	0.25 (0.18–0.33)	256 ± 36	21.0 ± 2.6	0.24 (0.18–0.31)
Urban/Rural Classificationd
Urban	520 ± 52	88.6 ± 3.0	NR	585 ± 56	85.4 ± 3.1	NR
Rural	67 ± 19	11.4 ± 3.0	NR	100 ± 23	14.6 ± 3.1	NR
Primary Payere
Medicare	91 ± 22	9.8 ± 2.3	NR	257 ± 36	21.1 ± 2.6	NR
Medicaid	271 ± 39	29.2 ± 3.4	NR	361 ± 45	29.7 ± 3.0	NR
Private (including HMO)	430 ± 47	46.3 ± 3.6	NR	350 ± 42	28.8 ± 2.9	NR

Abbreviations: CI, confidence interval; HMO, Health Maintenance Organization; NIS, National (Nationwide) Inpatient Sample; NR, not reported; RBF, rat-bite fever; RBI, rat-bite injury; SE, standard error.

^aWeighted number (SE) and associated percentages and rates are not presented for demographic strata with ≤10 unweighted hospitalizations. Rate (95% CI) is not presented where the relative SE > 0.30; these are represented by a dash. Weighted numbers may not add up to the total due to missing data.

^bAverage-annual hospitalization rate expressed per 1 000 000 persons.

^c“Asian/Pacific Islander”, “American Indian/Alaska Native”, and “Other” races as defined in NIS are collapsed into the Other category presented above due to small numbers. Rate for “Other” race is not reported due to no comparable denominator.

^dUrban/Rural Classification of patient’s residence is only available starting in 2007. Therefore, these numbers are out of a total of 605 (SE = 57) RBF hospitalizations and 705 (SE = 61) RBI hospitalizations. Rates are not reported due to no available denominator in the bridged-race population estimates.

^eOther primary payers analyzed but not reported in this table include “self-pay”, “no charge”, and other. No charge and other had too small of numbers to report for RBF, and no charge had too small of numbers to report for RBI. Rates are not reported due to no available denominator in the bridged-race population estimates.

We estimated 1216 (SE = 81) RBI hospitalizations for the study period, with an average-annual rate of 0.27 per 1 000 000 persons (95% CI, 0.23–0.30). The highest RBI hospitalization rates occurred among females, individuals ≥60 years of age, persons of black race, and persons in the Northeast (Table 2). We found no distinct primary payment mechanism, although approximately 60% of RBI hospitalizations were paid with Medicaid or private insurance. There were 107 (SE = 23) hospitalizations with both RBI and RBF diagnoses (8.8% [SE = 1.8%] of RBI hospitalizations).

The median LOS for RBF across the 15 years was 3.6 days (SE = 0.3). The median LOS for RBF was shortest for those 20–39 years of age (3.1 days; SE = 0.5) and longest for those ≥60 years of age at 7.5 days (SE = 2.0). Across age groups, the LOS for RBF was longer than for RBI.

Kids’ Inpatient Database

We estimated 27 (SE = 7), 31 (SE = 7), 38 (SE = 9), and 35 (SE = 9) RBF hospitalizations in children and young adults 0–19 years of age in 2003, 2006, 2009, and 2012, respectively. Concurrently, there were 21 (SE = 7), 35 (SE = 8), 22 (SE = 6), and 20 (SE = 6) RBI hospitalizations reported. Annual RBF hospitalization rates were lowest in 2003 (0.33 per 1 000 000 US children and young adults 0–19 years of age; 95% CI, 0.10–0.50) and highest in 2009 (0.46; 95% CI, 0.20–0.70). The RBI hospitalization rate was lowest in 2012 (0.25; 95% CI, 0.10–0.40) and highest in 2006 (0.42; 95% CI, 0.20–0.60).

State Inpatient Database

In SIDs, we identified 484 RBF hospitalizations with an average-annual rate of 0.21 per 1 000 000 person-years during 2001–2015 (Table 3). Almost half of the hospitalizations occurred in children and young adults (49.2%). Hospitalization rates were statistically significantly lower for age groups ≥20 years of age. Among those <20 years of age, 69.4% of hospitalizations occurred among individuals 5–14 years of age (data not shown). Hospitalizations occurred predominately among persons of white race, private insurance holders, and urban residents (reported 2007–2015). The median LOS was 4.0 days (interquartile range [IQR], 2.0–6.0); those 20–39 years of age had the shortest median stay (3.0 days; IQR, 2.0–5.0) and adults ≥ 60 had the longest (6.0; IQR, 3.0–11.0). Median charge per hospitalization was $15 077.50 (IQR, $7934.00–$29 340.50) (data not shown).

Table 3.

Number, Percentage, and Average-Annual Rates of Rat-Bite Fever and Rat-Bite Injury Hospitalizations by Demographic Attribute and Payment Type, SID, United States, 2001–2015^a

Characteristic	RBF			RBI
	No. (%)	Rateb	RR (95% CI)	No. (%)	Rateb	RR (95% CI)
Overall	484 (100)	0.21		725 (100)	0.32
Age Group (Years)
0–19	233 (49.2)	0.38	Reference	170 (24.3)	0.28	Reference
20–39	106 (22.4)	0.17	0.45 (0.36–0.56)	178 (25.4)	0.28	1.03 (0.83–1.27)
40–59	85 (17.9)	0.13	0.35 (0.28–0.45)	229 (32.7)	0.36	1.31 (1.07–1.59)
60+	50 (10.6)	0.12	0.31 (0.23–0.42)	124 (17.7)	0.29	1.06 (0.84–1.34)
Sex
Male	254 (53.5)	0.23	Reference	358 (49.8)	0.32	Reference
Female	221 (46.5)	0.19	0.84 (0.7–1.01)	361 (50.2)	0.31	0.97 (0.84–1.13)
Race/Ethnicity
White	318 (80.3)	0.21	Reference	392 (63.6)	0.26	Reference
Black	20 (5.1)	0.08	0.35 (0.22–0.55)	87 (14.1)	0.33	1.24 (0.99–1.57)
Hispanic	37 (9.3)	0.10	0.46 (0.33–0.64)	92 (14.9)	0.24	0.92 (0.74–1.16)
Otherc	21 (5.3)	–	–	459 (7.3)	–	–
Urban/Rural Classificationd
Urban	248 (82.7)	NR	NR	352 (85.0)	NR	NR
Rural	36 (12.0)	NR	NR	46 (11.0)	NR	NR
Primary Payere
Medicare	51 (10.5)	NR	NR	139 (19.2)	NR	NR
Medicaid	131 (27.1)	NR	NR	229 (31.7)	NR	NR
Private (including HMO)	226 (46.7)	NR	NR	199 (27.5)	NR	NR
Self-pay	45 (9.3)	NR	NR	78 (10.8)	NR	NR

Abbreviations: CI, confidence interval; HMO, health maintenance organization; NR, not reported; RBF, rat-bite fever; RBI, rat-bite injury; RR, rate ratio; SID, State Inpatient Database.

^aNumbers, percentage, and rates are not presented for demographic strata with ≤10 hospitalizations and are represented by a dash. Numbers may not add up to the total due to missing data.

^bAverage annual hospitalization rate expressed per 1 000 000 person years. SID is complete data and therefore standard error was not calculated.

^c“Asian/Pacific Islander”, “American Indian/Alaska Native”, and “Other” race as defined in SID are collapsed into the Other category presented above due to small numbers. Rate for Other race is not reported due to no comparable denominator.

^dUrban/Rural Classification of patient’s residence is only available starting in 2007. Therefore, these numbers are out of a total of 300 RBF hospitalizations and 419 RBI hospitalizations, with 5.3% and 5.0% missing, respectively. Rates are not reported due to no available denominator in the bridged-race population estimates.

^eOther primary payers analyzed but not reported in this table include “no charge” and other. No charge had too small of numbers to report for RBF and RBI. Rates are not reported due to no available denominator in the bridged-race population estimates.

We identified 725 RBI hospitalizations in SIDs for an average-annual rate of 0.32 per 1 000 000 person-years during 2001–2015. The RBI hospitalization rate was highest in adults 40–59 years of age and persons of black race (Table 3). Rat-bite injury hospitalizations occurred more frequently among Medicaid or private insurance holders (59.2%) and urban residents (reported 2007–2015). California, New York, and Florida reported 45.4% of all RBI hospitalizations and 41.1% of all RBF hospitalizations. Fifty-nine hospitalizations had both RBI and RBF diagnoses (8.1% of RBI hospitalizations).

Nationwide Emergency Department Sample

A total of 1033 (SE = 220) RBF ED visits were reported in the United States from 2006 to 2015, with an average-annual rate of 0.33 per 1 000 000 persons (95% CI, 0.19–0.47) (Table 4). A majority occurred among individuals 0–19 years of age, in the South, and among Medicaid or private insurance holders. Emergency department visits in the South accounted for 54.9% (SE = 9.8) of encounters, but we could not calculate rates given the large SE. In addition, a majority of ED visits were among urban residents and 51.1% (SE = 10.8) resulted in admission to the same hospital. A total of 32 588 (SE = 594) RBI ED visits were reported during the same time period with an average-annual rate of 10.51 per 1 000 000 persons (95% CI, 10.13–10.88). The average-annual rate was highest among individuals 0–19 years of age and in the Northeast. A majority occurred among those 0–19 years of age, in the South, and among Medicaid holders. A majority of patients were urban residents and treated and released from the ED. There were 105 (SE = 22) ED visits with both RBI and RBF diagnoses (0.3% [SE = 0.1%] of RBI ED visits).

Table 4.

Weighted Number, Percentage, and Average-Annual Rates of Rat-Bite Fever and Rat-Bite Injury Emergency Department Visits by Demographic Attribute and Payment Type, NEDS, United States, 2006–2015^a

Characteristic	RBF			RBI
	Weighted No. ± SE	Percent ± SE	Rate (95% CI)b	Weighted No. ± SE	Percent ± SE	Rate (95% CI)b
Overall	1033 ± 220	100	0.33 (0.19–0.47)	32 588 ± 594	100	10.51 (10.13–10.88)
Age Group (Years)
0–19	552 ± 179	53.4 ± 6.9	-	12 585 ± 312	38.6 ± 0.7	15.23 (14.49–15.97)
20–39	239 ± 50	23.1 ± 3.9	0.29 (0.17–0.40)	9528 ± 250	29.2 ± 0.6	11.38 (10.79–11.96)
40–59	173 ± 29	16.7 ± 3.7	0.20 (0.14–0.27)	6778 ± 217	20.8 ± 0.5	7.95 (7.45–8.45)
60+	69 ± 21	6.7 ± 2.2	-	3698 ± 152	11.4 ± 0.4	6.32 (5.81–6.83)
Sex
Male	524 ± 142	50.7 ± 4.4	0.34 (0.16–0.53)	15 657 ± 349	48.1 ± 0.6	10.27 (9.82–10.71)
Female	509 ± 91	49.3 ± 4.4	0.32 (0.21–0.44)	16 917 ± 371	51.9 ± 0.6	10.73 (10.27–11.20)
Region
Northeast	122 ± 25	11.9 ± 3.3	0.22 (0.13–0.31)	7713 ± 404	23.7 ± 1.0	13.91 (12.48–15.34)
Midwest	178 ± 32	17.2 ± 4.5	0.27 (0.17–0.36)	4633 ± 183	14.2 ± 0.5	6.91 (6.38–7.45)
South	567 ± 212	54.9 ± 9.8	-	12 737 ± 323	39.1 ± 0.9	11.05 (10.50–11.60)
West	166 ± 41	16.0 ± 4.7	0.23 (0.12–0.34)	7505 ± 229	23.0 ± 0.7	10.37 (9.75–10.99)
Urban/Rural Classificationc
Urban	903 ± 214	87.7 ± 3.8	NR	26 993 ± 560	83.3 ± 0.6	NR
Rural	126 ± 36	12.3 ± 3.8	NR	5406 ± 189	16.7 ± 0.6	NR
Primary Payerc
Medicare	57 ± 16	5.6 ± 1.8	NR	3696 ± 146	11.4 ± 0.4	NR
Medicaid	398 ± 145	38.6 ± 6.5	NR	10 656 ± 298	32.8 ± 0.7	NR
Private (including HMO)	396 ± 54	38.3 ± 6.0	NR	9394 ± 270	28.9 ± 0.7	NR
Self-pay	139 ± 42	13.4 ± 2.2	NR	5837 ± 205	18.0 ± 0.5	NR
No charge	-	-	NR	236 ± 38	0.7 ± 0.1	NR
Other	-	-	NR	2651 ± 126	8.2 ± 0.4	NR
Outcomed
Treated and released	494 ± 211	47.8 ± 11.0	NR	31 007 ± 580	95.2 ± 0.3	NR
Admitted to the same hospital	528 ± 60	51.1 ± 10.8	NR	842 ± 67	2.6 ± 0.2	NR
Not admitted, destination unknown	-	-	NR	645 ± 90	2.0 ± 0.3	NR

Abbreviations: CI, confidence interval; ED, emergency department; HMO, Health Maintenance Organization; NEDS, Nationwide Emergency Department Sample; NR, not reported; RBF, rat-bite fever; RBI, rat-bite injury; SE, standard error.

^aWeighted number (SE) and associated percentage and rates are not presented for demographic strata with ≤10 unweighted ED visits. Rate (95% CI) is not presented where the relative SE > 0.30; these are represented by a dash. Weighted numbers may not add up to the total due to missing data.

^bAverage-annual ED visit rate expressed per 1 000 000 persons.

^cRates are not reported due to no available denominator in the bridged-race population estimates.

^dOther outcomes analyzed but not reported in this table include the following: “Transferred to another short-term hospital”, “Died in ED”, and “Discharged alive, destination unknown (but not admitted)”. Died in ED and Discharged alive, destination unknown (but not admitted) had too small of numbers to report or represented zero ED visits for RBF and RBI. Transferred to another short-term hospital had too small of numbers to report for RBF and was only true for a small percentage of RBI ED visits. Rates are not reported because there is no applicable denominator.

Indian Health Service, National Patient Information Reporting System

We found no RBF hospitalizations reported in IHS NPIRS from 2001 to 2015. There were 66 RBF ED and outpatient clinic visits during this period, with an average-annual rate of 3.00 per 1 000 000 AI/AN (Table 5). The rates of RBF ED and outpatient clinic visits were not statistically significantly different between age groups and sexes. However, the Southwest region had a lower rate than the Northern Plains West, Southern Plains, and West.

Table 5.

Number, Percentage, and Average-Annual Rates of Rat-Bite Fever and Rat-Bite Injury ED and Outpatient Visits in American Indian/Alaska Natives (AI/ANs) by Demographic Attribute, IHS, United States, 2001–2015^a

Characteristic	RBF			RBI
	No. (%)	Rateb	RR (95% CI)	No. (%)	Rateb	RR (95% CI)
Overall	66 (100)	3.00		416 (100)	18.89
Healthcare Settingc
EDd	8	0.41	NR	142	7.34	NR
Outpatient Clinic	58	2.63	NR	274	12.44	NR
Age Group (Years)
0–19	29 (43.9)	3.47	Reference	235 (56.5)	28.14	Reference
20–39	17 (25.8)	2.53	0.73 (0.40–1.32)	99 (23.8)	14.72	0.52 (0.41–0.66)
40–59	14 (21.2)	2.99	0.86 (0.46–1.63)	61 (14.7)	13.04	0.46 (0.35–0.61)
60+	6 (9.1)	2.65	0.76 (0.32–1.84)	21 (5.0)	9.27	0.33 (0.21–0.51)
Sex
Male	29 (43.9)	2.76	Reference	201 (48.3)	19.11	Reference
Female	37 (56.1)	3.22	1.17 (0.72–1.90)	215 (51.7)	18.69	0.98 (0.81–1.19)
IHS Region
Southwest	8 (12.1)	1.06	Reference	132 (31.7)	17.48	Reference
East	-	-	-	24 (5.8)	35.06	2.01 (1.30–3.10)
Northern Plains East	-	-	-	11 (2.6)	7.49	0.43 (0.23–0.79)
Northern Plains West	16 (24.2)	5.57	5.26 (2.25–12.28)	75 (18.0)	26.11	1.49 (1.12–1.98)
Southern Plains	19 (28.8)	3.99	3.76 (1.65–8.60)	100 (24.0)	21.00	1.2 (0.93–1.56)
West	16 (24.2)	6.00	5.66 (2.42–13.22)	45 (10.8)	16.86	0.96 (0.69–1.35)

Abbreviations: CI, confidence interval; ED, emergency department; IHS, Indian Health Service; NR, not reported; RBF, rat-bite fever; RBI, rat-bite injury; RR, rate ratio.

^aNumbers, percentage, and rates are not presented for demographic strata with <5 ED and outpatient visits and are represented by a dash.

^bAverage annual ED and outpatient visit rate expressed per 1 000 000 AI/AN. IHS is complete data and therefore standard error was not calculated.

^cRisk ratios cannot be calculated due to different denominators.

^dThe West region is not included in the denominator for the rate of ED visits.

Although there were too few RBI hospitalizations to report during 2001–2015, there were 416 reported RBI ED and outpatient clinic visits with an average-annual rate of 18.89 per 1 000 000 AI/AN. More than half of the 416 visits occurred among those 0–19 years of age (56.5%) with a statistically significantly higher rate in this age group than all others. Rat-bite injury visits and rates were similar between males and females. The Northern Plains West, Southern Plains, and Southwest regions accounted for almost 75% of RBI visits; however, the highest rates were in the East, followed by the Northern Plains West. None of the ED and outpatient clinic visits had both RBI and RBF diagnoses.

DISCUSSION

Rat-bite fever surveillance is not routinely conducted in the United States; therefore, rates are not available from traditional public health data sources. In addition, although RBI is a risk factor for acquiring RBF, there is not a reliable estimate of RBI burden. To estimate RBF and RBI incidence, we analyzed 5 national administrative datasets for rat-bite-related encounters in the United States. From NIS, which provides a nationwide sample of hospitalizations, we estimated that the average-annual RBF hospitalization rate during 2001–2015 was 0.20 per 1 000 000 population; the RBI hospitalization rate was 0.27 per 1 000 000 population. Our findings reinforce that RBF is rare, yet suggest it occurs more frequently than demonstrated in the literature [12].

We found a higher rate and greater proportion of RBF to be in children and young adults, which is consistent with previous findings that over 50% of RBF in the United States occurs in this subpopulation [9, 12, 30, 31]. The rate of RBF hospitalizations for each year in KID was consistently higher than the average-annual rate for the general US population in NIS, supporting that RBF hospitalization is more frequent among children and young adults [9].

Older adults (≥60 years of age) comprised a greater proportion of total hospitalizations than ED visits. This is perhaps due to greater risk for complications or poor outcomes in older adults [32]. Although older adults had longer and more expensive hospital stays, our findings show lower rates of RBF hospitalizations compared with children and young adults. Because we did not find evidence of a difference in RBI hospitalizations between older and younger age groups, we do not believe differential hospitalization rates can be explained by less rat exposure among older adults. However, there could be differences in exposure to wild versus pet rats that we cannot elucidate from these data. Rat owners have been shown to perceive pet rats as posing less of a health risk than wild rats [33]. This may contribute to varied health-seeking behaviors for rat bites from wild versus pet animals.

Hospitalization rates also differed between RBF and RBIs by race. Rates were similar among white individuals but differed significantly among black and Hispanic individuals. This discrepancy warrants further investigation. In contrast, the urban/rural proportions found from each dataset are consistent with the 2010 US Census, which found that 85% of the population lives in metropolitan areas [34].

Rat-bite fever ED visits occurred almost equally among patients with private insurance and Medicaid; the proportions were also similar for RBI ED visits and hospitalizations. However, RBF hospitalizations were most frequent among patients with private insurance. Medicaid recipients are predominantly persons of low socioeconomic status with disproportionate representation of minority racial/ethnic groups. This suggests that Medicaid recipients may not have equal access to hospital-based treatment for RBF. Given systemic barriers to healthcare in the United States, similar discrepancies between Medicaid ED visits and hospitalizations may be expected for other rare zoonoses with mild or varied clinical presentation. Further information is required to investigate socioeconomic barriers to care.

Approximately 10% of RBI hospitalizations and 0.3% of RBI ED visits also had a RBF diagnosis. There are unverified historical claims that RBF infections occur in 10% of rat bites; however, the proportion of ED visits with both bite and infection do not support these claims [12, 35–37]. The origin of the claim perhaps only identified severe infections that required hospitalization, which aligns with the proportion of hospitalizations with both rat bite and infection [36]. We caution that the proportion of rat bites resulting in RBF infection is a difficult statistic to estimate, given the incubation period for RBF and that the number of RBI encounters is likely an underestimate of all rat bites occurring in the United States. We were only able to calculate the proportion of RBF cases recorded on the same record as a RBI.

Half of the patients with ED visits for RBF were admitted to the same hospital for further evaluation. The high frequency of hospitalization after RBF ED visit may suggest that patients are presenting to the ED at a stage in disease progression beyond initial nonspecific symptoms or a rapid progression of disease. From the data available, we could not identify prior attempts to seek care. In contrast, RBI ED visits almost always resulted in release from the facility.

Finally, further investigation is required to determine socioecological risk factors resulting in greater RBF vulnerability among AI/AN populations. It is known that 27% of AI/ANs residing on tribal lands live in conditions of poverty [38]. Houses on reservations have a higher tendency to be overcrowded and closely clustered, with higher rates of substandard infrastructure than houses nationwide [39, 40]. Inadequate housing with potentially higher susceptibility to rodent infestation suggests that RBF may be a neglected bacterial disease within underserved populations.

Limitations

This investigation provides the first characterization of rat-bite-related encounters in the United States using nationally representative data. However, there are several limitations. The unit of analysis for HCUP and IHS datasets are healthcare encounters and not cases. Readmission and transfer between healthcare facilities are frequently reported during RBF illness, with potential for a single patient to be captured by multiple encounters [41, 42].

Despite this, rates from HCUP and IHS likely underestimate RBF and RBI incidence. Streptobacillus moniliformis infections can manifest as subclinical cases or with mild symptoms. Diagnosis can be difficult, and RBF may not be considered given its rarity. A proportion of patients may seek services from healthcare facilities not captured within HCUP, and patients may administer self-care for rat bites and fevers that do not develop complications.

Similarly, rates of RBF and RBI encounters in AI/AN do not capture AI/AN patients who do not use the IHS. The IHS serves approximately 2.3 million people, but it does not reach all 4.1 million individuals identifying as AI/AN in the US Census [43–45]. Therefore, encounter rates are not generalizable to AI/AN populations nationwide.

The ICD-9-CM and ICD-10-CM codes are based on clinical management and medical charting practices, with potential inaccuracies in diagnosis and coding. Discharge records may include presumptive RBF diagnoses without laboratory confirmation, and codes have been shown to have varying levels of sensitivity and specificity for other conditions [13, 46–48]. In this study, it is not possible to discern the sensitivity of the codes used, thus we may be over- or underestimating the RBF and RBI rates. In addition, exposure history, infectious disease specialist consultation, and confirmatory laboratory testing are not included within hospital discharge or outpatient records but are critical considerations for correct RBF diagnosis.

Analytic limitations resulted from the rarity of RBF. All sample datasets had multiple years of ≤10 unweighted hospitalizations/visits or relative SE > 0.30, creating challenges for longitudinal analysis. Age categories were collapsed into 20-year intervals to retain statistical power, prohibiting exploration of higher-risk subgroups. Estimates were too unstable to accurately calculate 95% CIs for RRs; therefore, RRs and corresponding CIs were not presented for sample data.

Recommendations/Future Research

More information is needed to understand the epidemiology of the disease, particularly types of rodent exposures, disparities in care, and laboratory diagnosis [49]. However, this study provides the groundwork for future epidemiological research of RBF by identifying groups with higher rates of infection. Meanwhile, communication materials targeted to the groups identified in this study and those who have contact with rodents may be useful to educate on the potential risk.

CONCLUSIONS

For the medical community, clinical recognition involves enhancing awareness of RBF and the implications of this disease among pediatrics. Rat-bite fever should be a differential diagnosis for patients with (1) symptoms of fever, nausea, vomiting, joint pain, rash, and septic arthritis and (2) history of known or suspected exposure to rodents, even for patients without distinct bite or scratch wounds. Social and occupational histories should be taken, including inquiry into pet ownership and contact with rodents or other animals that may have had close contact with rodents. Because of its rarity, clinical references should provide specimen collection and processing procedures optimal for S moniliformis isolation. Diagnostic research should focus on developing rapid molecular techniques, better isolation methods, and whole genome sequencing-based analyses [50, 51].

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Acknowledgments

This study was made possible by data partners that voluntarily provide hospitalization data to the Healthcare Cost and Utilization Project (https://www.hcup-us.ahrq.gov/db/hcupdatapartners.jsp). We thank Sharada Ramakrishnan and William Bower for contributions to the study and review of this manuscript.

Author contributions. P. A. K. contributed to project concept development, results interpretation, writing, critical review, and editing. M. K. P. contributed to project concept development, analysis, results interpretation, writing, critical review, and editing. S. M. S. contributed to project concept development, analysis, results interpretation, critical review, and editing. J. R. M. contributed to critical review and editing. J. M. contributed to data acquisition, critical review, and editing. R. M. T. contributed to project concept development, results interpretation, writing, critical review, and editing.

Disclaimer. The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention or the Indian Health Service.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.