Cat scratches are common enough to matter in a country with nearly 40 million households that keep cats. Most attention focuses on cat-scratch disease (CSD), an infection caused by Bartonella henselae. The statistics below show how often diagnosed disease appeared in U.S. health-care data, which groups had higher rates, how atypical illness differed from typical illness, and how often Bartonella was found in selected cat populations.
Contents
- The basic cat-scratch disease picture
- U.S. outpatient and inpatient incidence
- Age, geography, and seasonality
- Hospitalization and atypical disease
- What cat and flea data show
- Bartonella findings in selected cat populations
The basic cat-scratch disease picture
The CDC identifies Bartonella henselae as the cause of cat-scratch disease. People usually develop the infection after a scratch from a domestic or feral cat, with kittens highlighted as a particularly common source. Bites from infected cats can also transmit the bacteria. Transmission can occur when a scratch is contaminated with flea feces, and the CDC notes that CSD can occur anywhere cats and fleas are found together.
The exposure does not need to look dramatic. The CDC says cat bites and scratches can spread germs even when a wound does not appear deep or serious. It also reports that about 20% to 80% of reported cat bites and scratches become infected. That range describes reported bites and scratches generally, rather than the share that become cat-scratch disease, so it should not be read as a CSD rate.
Rough play is one behavioral factor the CDC connects with exposure: playing roughly with cats and kittens increases the chance of scratches and bites. People with weakened immune systems face a higher risk of severe cat-scratch disease. Tender, enlarged lymph nodes typically develop 1 to 3 weeks after exposure. Rare infections can involve the eye, liver, spleen, brain, bones, or heart valves.
U.S. outpatient and inpatient incidence
A large U.S. claims study titled “Cat-Scratch Disease in the United States, 2005–2013” identified 13,273 CSD patients during the nine-year study period. The total included 12,735 outpatient cases and 538 inpatient cases. The analysis covered 280,522,578 person-years, with median MarketScan enrollment of 39,970,145 people per year. Annual enrollment ranged from 16,159,068 to 53,131,420 people.
The study’s average annual outpatient incidence was 4.5 cases per 100,000 population. Average annual inpatient incidence was 0.19 admissions per 100,000 population. These measures describe diagnosed cases and admissions captured in the claims study; they are not a count of every scratch or every infection in the community.
| Measure from the 2005–2013 claims study | Result |
|---|---|
| Identified CSD patients | 13,273 |
| Outpatient cases | 12,735 |
| Inpatient cases | 538 |
| Average annual outpatient incidence | 4.5 per 100,000 |
| Average annual inpatient incidence | 0.19 per 100,000 |
| Study period | 2005–2013 |
Outpatient incidence was highest at 5.7 cases per 100,000 in 2005 and fell to 4.0 per 100,000 in 2013. Inpatient incidence reached its highest reported level in the study at 0.22 admissions per 100,000 in 2008. The outpatient and inpatient series therefore describe different parts of the disease burden: outpatient diagnoses were much more frequent, while inpatient admissions represented a smaller, more severe subset.
For Americans younger than 65, the study estimated about 12,500 CSD diagnoses each year and an annual incidence of 4.7 cases per 100,000 people. That estimate is specifically for the under-65 population and should not be generalized to all ages without qualification.
Age, geography, and seasonality
Children 5 to 9 years old had the highest rates of both outpatient diagnoses and inpatient admissions in the 2005–2013 claims study. This age finding is consistent with the CDC’s warning that rough play with cats and kittens increases opportunities for scratches and bites, although the claims study and the CDC guidance answer different questions.
The strongest regional incidence levels in the claims analysis were reported for three Census divisions. Incidence in the West South Central, East South Central, and South Atlantic divisions was 6.1 to 6.4 cases per 100,000 population. By comparison, the Mountain division had an incidence of 2.2 cases per 100,000. The figures are division-level results from the study period, not current state-level estimates.
The South Atlantic division accounted for 26.3% of overall cases, while the West South Central division accounted for 19.7%. A region’s share of all cases is not the same as its incidence rate: the share depends partly on the size of the population included in the division.
An earlier analysis, “Cat scratch disease in the United States: an analysis of three national databases,” reported that about 60% of case-patients were discharged from September through January. The same analysis found that 55% of case-patients were 18 years old or younger and that males accounted for 60% of cases. These figures come from that earlier national-database analysis and should not be merged with the later 2005–2013 claims estimates as though they were one dataset.
Hospitalization and atypical disease
The earlier national-database analysis estimated hospitalized CSD incidence at 0.77 to 0.86 per 100,000 population per year. It also reported an ambulatory-patient incidence of 9.3 per 100,000 population per year and estimated annual health-care costs above $12 million. Those estimates use a different analysis and case definition from the 2005–2013 MarketScan study, which is why their rates are not directly interchangeable.
Connecticut surveillance for 1992–1993 reported that 11% of all CSD case-patients were hospitalized. This is a specific surveillance result from Connecticut and that period, not a universal hospitalization rate.
A separate U.S. claims study, “Atypical Manifestations of Cat-Scratch Disease, United States, 2005–2014,” included 14,600 typical cases and 224 atypical cases. Atypical disease represented 1.5% of all CSD in that study. Although atypical cases were a small share, their hospitalization risk ratio was 8.77.
The same study reported that 38.2% of typical cases were male, compared with 42.0% of atypical cases. A related report on atypical cat-scratch disease presentations found that 52% of atypical cases had no lymphadenopathy. It reported osteomyelitis in 31%, hepatic and/or splenic microabscesses in 31%, both osteomyelitis and hepatic/splenic involvement in 10.5%, and neuroretinitis in 5.2%.
These atypical-presentation percentages describe selected atypical cases, not all people who receive a CSD diagnosis. The distinction matters: a rare subgroup can have a substantially higher hospitalization risk without making severe complications typical for the overall patient population.
What cat and flea data show
The CDC’s transmission description places cats, fleas, scratches, bites, and flea feces in the same chain of exposure. The available cat studies add a different perspective: they measure Bartonella detection in selected animal populations, rather than human CSD incidence after a scratch.
The results vary considerably by location, cat age, and study method. That variation is important when interpreting any single prevalence number. A positive result in a cat population does not mean that every scratch transmits bacteria, and it does not provide a direct probability that a person will develop CSD after a particular scratch.
Bartonella findings in selected cat populations
In “Bartonella Infection among Cats Adopted from a San Francisco Shelter, Revisited,” 115 of 351 shelter cats were Bartonella-positive, a rate of 32.8%. Within the same shelter study, 50 kittens were culture-positive, or 25.9%, while 65 young adults were culture-positive, or 41.1%.
A Berlin study titled “Genetic Variability and Prevalence of Bartonella henselae in Cats in Berlin, Germany” reported B. henselae prevalence of 1% in pet cats and 18.7% in stray cats. In that same study, B. clarridgeiae prevalence was 0% in pet cats and 1% in stray cats. Because the study examined Berlin cats and separated pet from stray animals, these figures should remain tied to that geography and population.
| Selected cat population or finding | Reported result |
|---|---|
| San Francisco shelter cats, Bartonella-positive | 115 of 351; 32.8% |
| San Francisco shelter kittens, culture-positive | 25.9% |
| San Francisco shelter young adults, culture-positive | 41.1% |
| Berlin pet cats, B. henselae prevalence | 1% |
| Berlin stray cats, B. henselae prevalence | 18.7% |
| Berlin pet cats, B. clarridgeiae prevalence | 0% |
| Berlin stray cats, B. clarridgeiae prevalence | 1% |
Earlier feline bacteremia reports cited in “Immunofluorescent Detection of Intraerythrocytic Bartonella henselae in Naturally Infected Cats” found prevalence ranging from 22% to 59% across reports. Those reports also found that infected red blood cells ranged from 1% to 20%. In one infected-cat sample, 3% to 8% of erythrocytes were infected, with a mean of 5%.
At that 5% mean infection level, the authors estimated about 2.5 × 10^8 infected erythrocytes per milliliter of blood. The same study cultured only 10^5 CFU/ml from blood, implying that roughly 1 in 2,500 bacteria was cultivable under the reported conditions. This laboratory finding concerns detection and cultivation in feline blood; it is not a human transmission rate or a measure of how often scratches become infected.
Taken together, the statistics show why “cat scratch statistics” needs careful definition. Scratch and bite infections are broader than cat-scratch disease, human diagnoses are different from cat prevalence, and typical CSD is different from the smaller atypical group. The most useful comparison keeps the source, geography, measurement period, population, and outcome attached to every number.