Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Cat Scratch Disease Update


Cat-scratch disease has never been foremost in Bandolier's thoughts. For one thing it is relatively uncommon, is usually seen in children, and has a relatively straightforward diagnosis and treatment. There are a few hostages to fortune there, but that's the way it is oftentimes. Cat-scratch disease in older people wasn't even on the radar, but reports from a new study from Israel suggests that it has a different presentation than in younger people [1,2].


This was a surveillance study began in Israel in 1991, and involved clinicians sending specimens from patients suspected to have cat-scratch disease to a single laboratory. If they met case definition criteria, they were enrolled into the study and demographic and clinical data collected prospectively with a structured questionnaire, supplemented by medical records if necessary.

A case was defined as at least one positive test, including PCR, antibodies, or culture, with a compatible clinical syndrome. Typical disease was defined as regional lymph node involvement without ocular, cutaneous, or visceral problems. Atypical included one of several problems, including ocular, encephalitis, endocarditis, fever of unknown origin, erythema nodosum, hepatitis, splenitis, or osteomyelitis.


The group of 846 patients with confirmed cat-scratch disease formed the cohort. Typical disease was present in 85%. Overwhelmingly patients were young, with 80% of cases in under 40s, and 70% in under 30s. The average age was 18 years, with the highest number of cases between four and 14 years, and a smaller peak between 22 and 28 years (Figure 1, for age to 50 years).

Figure 1: Age distribution of presenting cases, up to age 50 years

Cat scratch disease was diagnosed in 52 patients at least 60 years or older [1]. Their presentation was different (Table 1). There were generally more women, but fewer with lymph node involvement. More of the older patients had atypical presentation, fever of unknown origin, encephalitis, and especially endocarditis. Diagnosis took longer in the over 60s, and fewer received antibiotics.

Table 1: Presentation in older patients

Percent with feature
<60 years
≥60 years
Odds ratio
(95% CI)
Lymph node involvement
0.2 (0.1 to 0.4)
General malaise
2.3 (1.2 to 4.3)
Female sex
2.0 (1.1 to 3.5)
Atypical presentation
3.1 (1.7 to 5.7)
Fever of unknown origin
7.3 (2.2 to 25)
6.3 (1.2 to 33)
62 (12 to 300)

A second report of the same study [2] identified 24 patients with arthropathy, not usually associated with cat scratch disease. These patients were aged between 23 and 54 years, and were more likely (5/24; 21%) to have erythema nodosum than in patients without arthropathy (2%).

These patients with arthropathy had frequent involvement of knee, wrist, or ankle. All had regional lymph node involvement, and in 22/24 arthropathy developed within a week of appearance of lymph node involvement.

Average duration of arthropathy was 13 weeks, but that encompassed the enormous range of one to 240 weeks. Nineteen (74%) recovered, and in these recovery came within 24 weeks (median six weeks). In the other five patients the disease became chronic, persisting for 16-53 months at the latest follow up.


It is comforting that we can continue to learn using good quality observation. Diagnosing cat scratch disease in older people may be made easier by the increasing availability of genetic methods like PCR, especially of tissue for the causative agent, bartonella henselae. An interesting, though not extensive, study of patients with lymph node involvement [3] tells us that PCR is 100% specific – meaning that a negative test is a rule out. Using PCR, and the starting point of superficial lymph node involvement in one isolated area, the paper has what looks to be a useful, if untested, diagnostic algorithm.


  1. R Bewn-Ami et al. Cat-scratch disease in elderly patients. Clinical Infectious Diseases 2005 41: 969-974.
  2. M Giladi et al. Cat-scratch disease – associated arthropathy. Arthritis & Rheumatism 2005 52: 3611-3617. 3 Y Hansmann et al. Diagnosis of cat scratch disease with detection of bartonella henselae by PCR: a study of patients with lymph node involvement. Journal of Clinical Microbiology 2005 43: 3800-3806.

previous or next story