Adverse local reactions from accidental BCG overdose in infantsJ M Puliyel, registrar in paediatrics,a A Hughes, hospital manager,a M L Chiswick, professor of paediatrics,a M Z Mughal, consultant paediatrician a
a Department of Paediatrics, Saint Mary's Hospital, Manchester M13 0JH
Correspondence to: Dr Mughal.
Local reactions to BCG vaccine depend on the administration technique, the dose, and the type of BCG preparation.1 We report on infants who were accidentally vaccinated intradermally with a percutaneous BCG preparation, receiving about five times the upper limit of the currently recommended intradermal dose of BCG.
All infants born at our hospital before November 1994 were routinely given intradermal BCG after written parental consent was obtained. A total of 857 infants were accidentally vaccinated intradermally with the percutaneous BCG (Evans E4981A and E4946B) between July and November 1994. After public announcement of the error, 556 of these infants attended special follow up clinics where they were examined for adverse reactions to BCG.
The median age at vaccination was 10 days (range 1-403 days). Infants were seen at the clinic a median of 65 (7 to 139) days after vaccination. A total of 61 infants (11.0%) had adverse local reactions. Forty eight infants (8.6%) had axillary lymphadopathy; one had an axillary lymph node >20 mm in diameter. Six infants had papules >10 mm diameter, and another six had ulcers >10 mm diameter. In one infant an abscess at the injection site was aspirated by needle.
One infant who received this BCG at 6 weeks of age presented at 4 months with severe combined immune deficiency (Omenn syndrome). She was treated with anti-tuberculosis drugs until her death from pulmonary haemorrhage after a bone marrow transplant. A lung biopsy three days before her death did not show histological changes of disseminated BCG.
The definition of an adverse local reaction to BCG varies greatly. O'Brien et al consider axillary lymph nodes >20 mm or vaccination ulcer prolonged for more than six weeks to be mild complications, axillary abscess or fistula to be moderately severe complications, and disseminated BCG infection to be severe complications.2 By these criteria only one child in our study had a mild reaction.
In older children, a normal BCG ulcer should not exceed >10 mm in diameter and should heal within four weeks.3 Six of our infants (1.1%) had ulcers >10 mm diameter, but this may be an underestimate as the infants were examined at different times after their vaccination. The size of the BCG ulcer depends on the technique of vaccination as much as the dose. In one report 158 of 403 children vaccinated by a doctor developed adverse local reactions; this was attributed to faulty technique.3 The low adverse local reaction rate in our cohort may, despite the high dose of BCG used, reflect the experience and good intradermal vaccination technique of the two doctors who administered the vaccine.
Surprisingly, the patient with Omenn syndrome did not show evidence of disseminated BCG infection. This syndrome in its early stage is characterised by polyclonal proliferation of T lymphocytes, and we speculate that these T lymphocytes may have been capable of activating macrophages, thus preventing dissemination of BCG in this patient.
While human error was responsible for this "accident," it is important to note that the ampoules, packaging, and labelling of the Evans intradermal and percutaneous BCG preparations are deceptively similar. Distinctive labelling and packaging of the intradermal and the percutaneous BCG preparations would have helped to draw attention to their different potency.
We thank Ms A Fisher, Dr F N Bamford, Dr R Pumphery, Dr G Morgan, Professor RHD Boyd, and staff of the Central Manchester Healthcare and the Mancunian Trusts.
Conflict of interest: None.
Source of funding: None.