Home | Help | Search/Archive | Feedback | Table of Contents |
|
Does screening really reduce mortality?
EDITORWe were
rather non-plussed to read that the conclusion of the paper by Quinn et al on
screening for cervical cancer1
is not supported by their data, and we wonder whether so called
political correctness had anything to do with it. The statement
"800 deaths might have been prevented in 1997" is based on a
projected mortality of a completely arbitrarily (alas, not randomly)
selected part of a subset of graphs showing trends in mortality. The
opposite conclusion may be reached using the same graphs. For
example, in women aged 35-44 mortality fell from 10 per
100 000 to 5 per 100 000 in the period 1960 to
1975, and it should have approached zero by 1997 assuming
that the trend had continued. Similarly, with the same age groups as
in the original paper, in women aged 25-34 mortality fell from
2.5 per 100 000 to 1.1 per 100 000 in the period
1955 to 1965, so by 1997 it should have again approached
zero. Since the only new intervention has been screening, and
the mortality is excessive at 5 per 100 000, screening may
have caused up to 2900 extra deaths in 1997by the
same logic.
j.vaidya@ucl.ac.uk
Michael Baum
Department
of Surgery, Royal Free and University College Medical School, University College
London, London W1P 7LD
1. | Quinn M, Babb P, Jones J, Allen E, on behalf of the United Kingdom Association of Cancer Registries. Effect of screening on incidence of and mortality from cancer of cervix in England: evaluation based on routinely collected statistics. BMJ 1999; 318: 904-908[Abstract/Full Text]. (3 April.) |
Authors' reply
EDITORThe
conclusions in our paper are not based solely on the analysis of mortality. We
presented strong evidence that the introduction of national call and
recall and of incentive payments to general practitioners led to a
dramatic fall in the incidence of cervical cancer in women in all age
groups from 30 to 74 and in all regions of England. Other
evidence confirms the expected shift towards detection of earlier
stages of disease. There is no other plausible explanation for these
patterns. If women do not get cervical cancer, they will not die from
it. In addition, it has been recognised for over 30 years that
mortality from cervical cancer shows very strong cohort trends
(reflecting those in incidence)1
and so Vaidya and Baum's simple extrapolation of age specific
trends is totally inappropriate. We extrapolated the cohort rates
for the relevant age groups. Our analysis and conclusions are
supported by a similar study in Scotland2 and by
the results from formal age period cohort models.3 We remain deeply concerned about the many well known problems with cervical
screening which we mentioned in our paper: cervical cancer is a
comparatively rare disease and its natural course is not well
understood; the smear test has both low sensitivity and low
specificity; many tests are technically unsatisfactory and the
proportion of such tests varies widely across the country; the mix of
three and five year screening intervals is inequitable; too many
smear tests are opportunistic; and the programme costs four times as
much as breast screening. Nevertheless, there is now conclusive
evidence that cervical screening has markedly reduced both incidence
and mortality.
Study shows importance of centralised organisation in
screening
EDITORThe
paper by Quinn et al reporting the effects of screening on incidence of and
mortality from cervical cancer in England1
highlighted the characteristics of successful programmes elsewhere 2 3
and showed that the national screening programme had been
effective. The situation in Hong Kong, where there is no systematic population based
cervical screening programme, shows the importance of central
organisation. Hong Kong is a generally affluent community with a
better health profile than most developed countries. Infant mortality
is low (4.6 per 1000 live births in 1995, compared with
6.2 in the United Kingdom), and life expectancy is high
(81.5 years at birth for women, compared with 79.4 years in
the United Kingdom). Women in Hong Kong are at lower risk of
developing many common cancers, such as those of the breast and lung,
than are their counterparts in most Western countries yet the reverse
is true for cervical cancer.4 The figure shows the trend in the incidence of and mortality from cervical
cancer standardised to the European standard population (for age
bands of five years). Although incidence has reduced gradually over
time, it has not fallen dramatically as in the United Kingdom after
organised screening achieved a coverage greater than 70%, and the
death rate has changed little. The standardised incidence of
16.9 per 100 000 for invasive cancer in 1994 was
higher than the baseline rates of disease before organised screening
started in the United Kingdom. Cervical cancer is the fourth most
common newly diagnosed cancer and accounts for 4% of deaths from
cancer in local women, compared with 2% in the United Kingdom.
One of us (PA) recently found that 56% of nearly 1800 women aged between
20 and 75 in Hong Kong had never had a cervical screening
test.5
Coverage was lowest among older women (72% of women over 50 had
never been screened) and those in the lower socioeconomic groups.
Less than a quarter of all women were screened regularly, and these
were generally screened yearly or more often.
The current screening system in Hong Kong is therefore inequitable, wastes
resources, and results in avoidable cases of cervical cancer. It may
also cause unnecessary harm by overscreening women at lower risk. The
study by Quinn et al provides further support for centralised
organisation in any screening system and is a message that should not
be ignored by any country with a developed health care
system.
P
J Babb
J Jones
Office for National Statistics,
London SW1V 2QQ
1.
Hill GB, Adelstein AM. Cohort mortality from carcinoma of
the cervix. Lancet 1967; ii: 605-606.
2.
Walker JJ, Brewster D, Gould A, Raab GM. Trends in
incidence of and mortality from invasive cancer of the uterine cervix in
Scotland (1975-1994). Public Health 1998; 112: 373-378[Medline].
3.
Sasieni P, Adams J. Effect of screening on cervical cancer
mortality in England and Wales: analysis of trends with an age period
cohort model. BMJ 1999; 318: 1244-1245[Full
Text]. (8 May.)
View larger
version (18K):
[in a new window]
Age standardised incidence of invasive
cervical cancer and mortality from cervical cancer, Hong Kong,
1982-7
Sarah McGhee
Anthony Hedley
Department of Community
Medicine, University of Hong Kong, Patrick Manson Building South Wing,
7 Sassoon Road, Pokfulam, Hong Kong
1.
Quinn M, Babb P, Jones J, Allen E, on behalf of the United
Kingdom Association of Cancer Registries. Effect of screening on incidence
of and mortality from cancer of cervix in England: evaluation based on
routinely collected statistics. BMJ 1999; 318: 904-908[Abstract/Full
Text]. (3 April.)
2.
Laara E, Day N, Hakama M. Trends in mortality from cervical
cancer in the Nordic countries; association with organised screening
programs. Lancet 1987; i: 1247-1249.
3.
ICRF Coordinating Committee on Cervical Screening.
Organisation of a programme for cervical cancer screening. BMJ
1984; 289: 894-895[Medline].
4.
Adab P, Hedley AJ. Preventing avoidable death: the case of
cervical cancer in Hong Kong. Hong Kong Med J 1997; 3: 427-432[Medline].
5.
Adab P. Screening for cervical cancer in Hong Kong
[abstract]. Fifth Hong Kong international cancer congress, Hong Kong ,
February 1998.
© BMJ
1999
|
Read all Rapid Responses
Home | Help | Search/Archive | Feedback | Table of Contents |