Visit to Public Health Laboratory
Service Headquarters, Colindale, 23 October 1997
1. Three members of the Sub-Committee,
with staff, paid a second visit to PHLS at Colindale.
2. Dr David Livermore showed us
the Antibiotic Reference Unit, where isolates referred
from hospitals are tested for susceptibility or resistance. Usually
only unusual or problematic isolates are referred (except for
TB and salmonella); hence the selectivity of surveillance. We
were shown four different kinds of test.
(i) The pathogen is cultured
on an agar plate, and discs of antibacterial agent are then placed
on the plate. A clear circle round the disc indicates susceptibility.
This test takes two days; it is cheap (10p per disc); it is only
(ii) The pathogen is cultured on
an agar plate, then the antibacterial agent is introduced on a
strip marked with an ascending scale of concentration. This test
is more expensive (£1.50 per strip) and no faster, but quantifies
the "minimum inhibitory concentration" (MIC).
(iii) The pathogen is cultured on
a series of plates containing an antibacterial agent in increasing
concentrations, to identify the MIC. We saw a computer-assisted
technique for reading the plates, but this still relied heavily
on human hand and eye.
(iv) The pathogen is tested for
the presence of genes associated with resistance, using the polymerase
chain reaction (PCR). This test is more expensive than tests (i)(iii)
(£5 per test), and only slightly faster; and each test
will only show the presence or absence of a single known gene.
We saw simple kits in use; automated systems are under commercial
3. We asked about the application
of digital imaging to these tests; we were told that automation
in microbiology is lagging behind other laboratory disciplines.
4. Dr Barry Cookson showed us the
Laboratory of Hospital Infection, and discussed his survey
of infection control teams and MRSA in 1995. He noted that antibiotic
policies are best informed by setting information about local
prescribing patterns against information about local resistance
patterns: paradoxically, in hospitals, microbiological information
tends to be good while prescribing information tends to be poor,
while in general practice the situation is reversed. He found
only three hospitals where antibiotics policies were firmly based
on the findings of local surveillance; and he pointed to King's
College Hospital as an unusual example of success in using local
infection data to negotiate for more resources for infection control.
In general practice, PHLS aims to set up a network of 50 sentinel
practices; but it is hard to persuade a GP to devote time and
money to testing, until empirical treatment has been tried and
5. Dr Cookson mentioned that, on
a surgical ward, the rate of patients colonised with SA who develop
infection may be one in four. He commented on the very low incidence
of MRSA in the Netherlands: infection control is stringent; and,
because of the low incidence, stringent control is affordable.
He also commented on the British National Formulary: much good
information, but presented in an "unfriendly" fashion.
He noted the particular difficulty, in surveillance of MRSA, of
establishing the source of infection.
6. Dr Bernard Rowe showed us the
Laboratory of Enteric Pathogens. This is the national reference
laboratory for food-borne pathogens, and is linked to other reference
laboratories in Europe via ENTERNET, funded by the EU; it is also
a WHO Collaborating Centre for resistant strains. The Laboratory
receives 95 per cent of all human salmonellas isolated in
the UK (except Scotland) and the Republic of Ireland, amounting
to 30,000 isolates per year; it receives another 20,000 isolates
per year of E. coli and other pathogens; it tests
up to 300 isolates per day, and has a database going back 30 years.
(Salmonella in animals is notifiable to the Central Veterinary
Laboratory under the Zoonoses Regulations.)
7. Dr Rowe insisted that the problem
of resistant salmonella was due to the use of antimicrobials in
animal husbandry as prophylactics (treating healthy animals) or
"metaphylactics" (treating healthy animals which have
been exposed to disease). Resistance in salmonella does not normally
threaten the individual patient, because in most cases recovery
takes place without treatment; but it increases the survival and
spread of salmonella in animals and therefore the overall level
of human food poisoning. There was no suggestion that genuinely
sick animals should not be treated; and, since the Swann Report
of 1969, the use as animal growth-promoters of antibiotics used,
or related to those used, in human medicine has been prohibited
in the UK.
8. We heard about the case of apramycin,
a veterinary antibiotic for calves. It is now known that use of
apramycin gives rise to resistance to gentamicin, an important
clinical antibiotic. Apparently, the industry was aware of this,
but waited until it was established by published research before
voluntarily withdrawing apramycin.
9. We also heard about the case
of enrofloxacin, licensed by MAFF for animal use in 1993, in the
face of scientific advice that it would give rise to resistance
to ciprofloxacin. Since 1994, Salmonella typhimurium DT 104
has acquired chromosomal resistance to ciprofloxacin. It was suggested
that this was a case of "agency capture", and that an
independent Food Standards Agency might have acted differently.
10. Dr John Watson told us about
the Communicable Disease Surveillance Centre, of whose
respiratory diseases group he is the head. CDSC has recently taken
over from OPCS the receipt of statutory notifications of infectious
diseases. Dr Watson explained the dual purpose of surveillance:
to advise individual doctors how to treat individual cases or
groups of cases; and to inform general treatment guidelines.
11. Dr Watson spoke mostly about
TB. UK surveillance of TB is comprehensive, and co-ordinated since
1993 by "MYCOBNET". A detailed survey is conducted every
5 years. Compared with salmonella, TB produces fewer isolates
(15,000 in 4 years, compared with 30,000 in one); and isolates
seen by the reference laboratory represent a much higher proportion
of all cases. (In TB 60 per cent of cases are identified
microbiologically, the rest clinically or by X-ray; in salmonella
only about one per cent of cases are identified at all.)
In the UK, TB resistance to isoniazid is stable at 6 per
cent, slightly lower than in the USA; resistance to isoniazid
and rifampicin ("multi-drug resistance", MDR) is low
(1-2 per cent) but clearly rising.