Table 3.
Risk factors for acquiring recurrent bacterial meningitis, Sweden 1997–2014. Multivariate survival model including age, gender and variables with a univariate p‐value <0.1 and a relative standard error <0.3 as estimated by using logistic regression with adjustment for calendar time and follow‐up time a . The adjusted relative risk for calendar time in 1‐year intervals (continuous variable) was 1.1 (1.1–1.2), p <0.001 b . To further minimise potential bias due to different versions of the International Classification of Diseases (ICD), the analysis was restricted to the period after the introduction of ICD10 in the year 1997
Risk factors of recurrent bacterial meningitis | Adj. relative risk c (95 % confidence interval) | p‐Value |
---|---|---|
Age categories and gender | ||
Age 0–4 years | (ref) | (ref) |
Age 5–17 years | 0.8 (0.3–1.8) | 0.53 |
Age 18–64 years | 0.9 (0.5–1.5) | 0.67 |
Age 65 years or more | 0.8 (0.4–1.4) | 0.43 |
Male gender | 1.1 (0.8–1.6) | 0.59 |
Lesions of the central nervous system d | ||
Intracranial shunt present | 2.8 (1.4–5.5) | 0.002 |
Cochlear implant present | 8.0 (4.3–15.1) | <0.001 |
Basilar skull fracture | 3.7 (1.6–8.3) | 0.002 |
Meningeal tumour e | 4.8 (1.4–16.2) | 0.01 |
Ischemic cerebrovascular | 2.9 (1.5–5.7) | 0.001 |
Malignant neoplasia | ||
Haematological malignancy | ||
Lymphoma | 2.4 (0.7–7.8) | 0.16 |
Leukaemia | 2.0 (0.6–6.9) | 0.30 |
Other malignant neoplasia | 1.3 (0.6–2.6) | 0.46 |
Comorbidity | ||
Moderate to severe liver disease | 2.8 (0.9–9.1) | 0.09 |
Primary immune deficiency | ||
Humoral deficiencies | 17.9 (6.5–49.4) | <0.001 |
Complement deficiencies | 7.2 (1.6–32.8) | 0.01 |
Other deficiencies | 3.1 (0.3–35.8) | 0.37 |
Monoclonal gammopathy of unknown significance | 5.8 (1.8–18.6) | 0.003 |
Only the first and second episode per patient were retained for this analysis.
Including all recurrent episodes, first time episodes excluded.
Relative risk estimated using negative binomial regression clustered on patients with the logarithm of time as an offset variable and robust variance estimation.
Neurosurgical, traumatic, neoplastic and cerebrovascular lesions of the central nervous system.
Including malignant and benign tumour forms.