20 Swedish National Fod Administration. Swedish Food ReguLlationis. Food
additives. Uppsala: Swedish Nationali Food Administration, 19X5.
21 Fine DH, Ross R, Rooinhehler DlP, Silvcrgleid A, Song L. Formation in vivo
of volatile N-nitrosamines in man aftcr ingestion of cookcd hacon and
spinach. Nature 1977;265:753-5.
22 Osterdahl BG. Volatile nitrosamines in foods on the Swedish market and
estimation of their dails intake. FtoodAddit Contam 1988;5:587-95.
23 Wagner DA, Shuker D)EG, Bilmazes C, et al. Effect of v%itamin C and E.
on endogenotis synthesis of N-nitrosamino-acids in humans: precursor-
product studies with (' N) nitrate. Cancer Res 1985;45:6519-22.
24 Jenkins D)JA, Wolever 'rMs, Jenkins AL,
carbohydrate metabolism and diabetes. Mo/lAspectsMed 1987;9:97-1,12.
25 AMiettinen 0. Components of the crtude risk ratio. Am J Epidemiol 1972;96:
26 Klecka WR. Discriminant analysis. In: Nie NH, Hull CH, Jenkins JG,
laylor RH. Dietary fibre,
Steinhrenncr K, Bent DH, cds.Starisnical packageforih.social sclent(cs 2nid
ed. Ncw York: McGraw-Hill, 1975:434-47.
27 Miettinen 0. Estimahility anid cstimation in cast-referent studies. Am 7
28 Rakicten N, Rakicten ML, Nadkarni MV. Studics on the dia!sctogenic action
of streptozotocin. Cancer Chemoiheraps' Reports 1963;29:91 -8.
29 Scott FW, Daneman 1), Martin JM. Evidence for a critical rolc of diet in tlc
development of insulin-dependent diahetes incillitis. Diabeties Ris 1988;7:
30 Nerup J, Mandrup-PoulsenT, Molvig J, Helqvist 8, S
Mechanisms of pancreatic B-cell destructioin in type I diahetes. Dlsahts
31 'Iibblin G. A pipulation studv oii 50-year-old-men. An analsis ot the nuoi-
participating grotip. Acia Med St-and 1965;178:453-9.
'ogcnscn L, Egebergi.
(Accepted 6 March 1990)
Has the prevalence of asthma increased in children? Evidence
from the national study of health and growth 1973-86
P GJ Burney, S Chinn, RJ Rona
Objectives-To estimate changes in the preva-
lence of reported symptoms of respiratory disease
and reported diagnoses of asthma and bronchitis in
primary school children in England between -1973
Design-Mixed longitudinal survey.
primary schools in 22 areas.
Participants-15 000 Boys and 14156 girls each
studied at least once between 1973 and 1986.
Data collected-Whether, according to the parent
or guardian, the child had wheezed, wheezed on
most days or nights, or had attacks of bronchitis or
asthma in the past year.
Results-Within age groups trends in successive
annual cohorts showed an increasing prevalence of
asthma for each annual birth cohort (boys, 6-9%,
p<0001; girls, 12-8%, p<OOOl) and of wheeze
on most days or nights (boys, 4-3% per cohort,
p<0-001; girls, 6-1% per cohort, p<0001) and a
falling prevalence of bronchitis (boys, -4-7% per
cohort, p<0001; girls, -5-8% per cohort, p<0001).
There was a smaller increase in the prevalence of
wheeze whether or not it occurred on most days or
nights, and this increase was significant only among
the girls (boys, 1-0% per cohort, p>005; girls, 1-7%
per cohort, p<O0O5). Although the rate ofincrease of
"asthma" was greater than the rate of decrease in
"bronchitis," the baseline prevalence ofasthma was
much lower than that of bronchitis, and the total
proportion ofchildren with either diagnosis declined
slightly over the whole period. The main change was
an increase in the proportion of children whose
parents stated that they had persistent wheeze and
yet did not have a report of either "asthma" or
Conclusions-These results suggest that there has
been a true increase in morbidity that is not simply
due to changes in diagnostic fashion. The increase is
large enough to explain much ifnot all ofthe increase
in admission to hospital and mortality, and
underlines the importance ofan understanding ofthe
aetiology of asthma in tackling the causes of the
Mortality from asthma among 5-34 year olds
increased between the mid-1970s and the mid-1980s.'
This increase was the more surprising as mortality had
been falling in most conditions for which there was
effective prevention or treatment.2 At the same time
more children with asthma had been admitted to
hospital34 and consultations with general practitioners
for asthma virtually doubled between 1970-1 and
Each of these changes could be due to a change in
medical practice, including a change in the use of
diagnostic labels, or to a change in the prevalence ofthe
disease. There is some indirect evidence that the
prevalence of asthma may have been increasing, but
other evidence suggests that there has been no change.
Smith found an increase in the prevalence ofasthma in
Birmingham in -the 1960s,6 but this evidence was
collected at a time of great demographic change
when the city was being largely rebuilt. Wadsworth
compared those who had been studied in the 1946
national perinatal survey with their firstborn offspring
and showed that. the children had a threefold greater
chance of having been treated for asthma before their
fifth birthday,7 but this could have been explained by
differences in management or differences in diagnostic
practice. Hill- et al estimated that the prevalence of
.1-28% between 1985 and 1988 and suggested that
the large increase in estimated '.'asthma" was due
to a change in diagnostic labelling.8 More recently
Burr et al showed increases between 1973 and.1988 in
both symptoms and bronchial response to exercise in
12 year old children attending schools in Caerphilly.9
other hand, Hay and Higginbottam,
Anderson, and Hill et alhave all pointed to the lack of
any trend over time in the results from published
surveys reporting the prevalence of asthma.'0'2 As
Anderson points out, however, the interpretation of
this evidence is difficult because the methods used
in each survey are different and estimates of the
prevalence ofasthma are likely to be sensitive to these
differences.'3 Geographical variation in the prevalence
of asthma would also confound any estimate of trend
from these data.
that asthma may be an increasing
problem are not confined to the United Kingdom.
Upward trends in mortality have also been noted
-in New Zealand, France, Germany; Denmark, and
possibly the United States.4'-6 Upward trends in
hospital admissions have also been noted in New
Zealand and the United States.'7
This paper reports trends in the prevalence of
respiratory conditions, including both diagnoses and
symptoms, reported by the parents and guardians of
children in the-national study of health and growth
between 1973 and 1986. Estimates of the trends in
prevalence have been made for cohorts of children
living in England and born between 1961 and 1981.
Health Medicine, United
Medical and Dental
Schools ofGuy's and
St Thomas's Hospitals,
St Thomas's Hospital,
London SEI 7EH
P G J Burney, FFPHM, senior
lecturer in public health
S Chinn, MA, senior lecturer in
R J Rona, MFPHM, senior
lecturer in public health
BrMedJ 1990;300: 1306-10
questionnaire was retuirnled and infor
NUMberof English boys
TABLE III-m Numbers (°/o) of cJlildre:
Number in sample w ith
Occasiotial wx heeze
Persistcnt wxhccc btut J1o
attacks o) asthma or
Numbcr in samiiple with
Occasionail w heiez
Persistcnt w hcezc
Persistenit wsheeze but nlo
attacks olt asthnma or
Subjects and methods
The subjects were English children aged from 4 to
12 years who took part in the national study of health
and growth at least once between 1973 and 1986. In the
national study of health and growth, which is primarily
a surveillance study, primary schools in 22 areas in
England were selected in 1972 and visited annuallv
until 1982 and biennially thereafter. Children were
eligible to take part while attending a participating
school and as a result children have been included up to
eight times between 1972 and 1982 and up to four times
since then. The original design was described in detail
bv Chinn and Rona."9
Thirteen areas continued in the study throughout
the period with the same schools participating. Three
continued with one change ofschool and the remaining
six were replaced once, three in 1977 or 1978, and three
in 1982 or 1984.
Data were collected on respiratory conditions by
questionnaires with some variation in content. The
question, "Has he or she suffered from any of these
illnesses in the last twelve months? Asthma
." was asked from 1973 to 1976 about all
children who were not new entrants to the study and
from 1977 onwards about all children. The questions
"Does his or her chest ever sound wheezy or whistling?"
and, if yes, "Does he or she get this on most days or
nights?" were asked of all children followed up
between 1973 and 1976 and all children in 1977,
1982, 1984, and 1986. A positive answer to the first
question defined those with "occasional" wheeze, and a
positive answer to both questions defined those with
A trend in prevalence over time for any of the
svmptoms or conditions
because of the repeated data at different ages for most
of the children
the derivation of an appropriate
Clarke described a method that can be computed using
the CATMOD procedure in SAS.
each age group from 5 to 12 (defined by year of survey
less the year of birth) there were three categories for
each respiratory condition: absent, present, or missing
information. Children were divided by sex and into
cohorts by year of birth, and the data for each condition
were cross tabulated by age group. Some data were
available for 21 cohorts born in 1961 to 1981 (table I).
For each cohort and eachage group we found the
proportion (p) of those with the condition or symptoms
of those in the cohort for whom information was
obtained and calculated the logit (log,p/(1-p)) in
accordance with usual methods for the analysis of
The pooled linear trend within age
groups over cohorts with the appropriate standard
error was then calculated using CATMOD. Owing to
the very large size of the cross tabulation it was not
possible to perform the analysis in a single run of
SAS on the University of London computer centre's
Amdahl machine. To
asthma and bronchitis the cohorts were divided into
three groups of every thirdcohort,and the resulting
three trends were averaged. For wheeze, as there were
fewer data, two groups of everv other cohort were
Quadratic time effects were calculated to assess the
adequacy of the linear trends as summaries of the data.
The trend in logit prevalence was multiplied by 13 to
give an estimate of change over the 13 years of the
study; the antilogarithm was derived and multiplied by
100 to give an approximate percentage increase. A
95% confidence interval for the 13 years' change was
calculated as ± 1 96x 13 x standard error oftrend,and
the antilogarithms for the limits were derived and
multiplied by 100 to give a 95% confidence interval for
the percentage increase.
[ABLE I- Data availablefor each cohort
is easily calculated,
is not straightforward. Woolson and
For each child in
`wsheeze, A -asthma or bronlchitis.
*Year of data collection minus sear of birth.
and girls eligibleJor studv and percentage of total eligiblefJor whom
rinationz onz respirators! onditions obtained
overcome the difficulty for
tn with reportedsvmptomsand diagnoses in 1973
3 (26 9
*Ntumbers vary slightlv; mitxtnxLm gliscri. Percentagcs arc ot actual nulmbcr.
About 6500 children were eligible for follow up each
year from 1973 to 1976, and around 8000 (decreasing to
around 6600) were eligible for the study between
1977 and 1986 (table II). Questionnaires were returned
for between 860o and 91t% of the participants; most
gave information on respiratory conditions. Com-
parable data were not obtained from new entrants
from 1972 to 1976 and so the data from the 1972 survey
have been omitted and data on new entrants have been
excluded between 1973 to 1976. New entrants to the
VOLUME 300 19 MAY 1990
study during this period who were not followed up are
therefore totally excluded from this analysis. A total of
15000 boys and 14 156 girls were included at least
once. There was little difference in the amount of
information available on asthma or bronchitis or, in
surveys in which the information was available, for
occasional or persistent wheeze. In 1973, 95 3% of the
quetmionnaires were completed by the mother or
female guardian, 3 9% by the father or male guardian,
and 0-8% by others. In 1986 these figures were 90 4%,
9.4%, and 0 I% respectively.
Table III shows the prevalence ofreported symptoms
and diagnoses in 1973, the first year of the study.
Symptoms were generally more common in boys, and
occasional wheeze and a diagnosis of bronchitis were
clearly much less common in older children. Table IV
shows changes in the percentage prevalence of one of
these symptoms, persistent wheeze, in succeeding
surveys for boys and girls separately. A general trend
can be seen within each age group for rates to increase
in succeeding cohorts. The calculations in table V
confirm that these apparent increases are greater than
could be expected by chance.
TABLE IV- Prevalence (0°) ofpersistent wheeze
There was a definite, highly significant (p<0l001)
rate of increase in the prevalence of reported asthma
over this period, and this was significantly greater
in girls than boys. There was a smaller but also
highly significant rate of fall in the prevalence of
reported bronchitis over the same period, and this was
approximately equal in girls and boys. There was a
highly significant rate of increase in the prevalence of
persistent wheeze, which was estimated to be slightly
less than that in the prevalence of reported asthma in
boys and approximately equal in both sexes. The
estimated prevalence of occasional wheeze increased
this was only marginally
(p<005) in the girls. The quadratic time effects were
not significantly different from zero except for that for
asthma in girls (p<005); in this case it represented a
trend that was itself increasing over time.
When diagnoses of "asthma" and "bronchitis" are
combined there is a small fall in the prevalence of the
total over the period, which is significant among boys
(p<005) but not among girls. In both sexes there was a
significant increase in the number of children with
wheeze on most days or nights but without a diagnosis
of either "asthma" or "bronchitis" (p<0 001). This
change was greatest in boys.
This analysis shows a definite increase in the
prevalence of diagnosed asthma and of "wheeze on
most days and nights" in English primary school
children between 1973 and 1986 and provides the first
national estimate of the increasing prevalence of
asthma. It strongly suggests that local factors do not
explain earlier observations"9 and that the review
of surveys from different centres using different
techniques'''1' is misleading.
The initial areas were selected by stratified random
sampling of employment exchange areas, and local
medical and education authorities selected the schools
within the chosen areas. The stratification of the
employment exchange areas took account of the level
ofunemployment, uptake offree school meals, and the
proportion of children leaving school at age
Though proportionally more areas were selected from
the poorer areas, the distribution of the children by
social class at the start of the study in 1972 was very
close to the equivalent national figures for England and
Wales at the 1971 census.2' Estimates of change based
on 22 different areas covering diverse parts ofEngland
are unlikely to be attributable to local changes in the
composition of the population.
The methods used have been the same throughout
the period. The same questions on respiratory health
had been asked throughout the study, though questions
on wheeze were not asked in all years. In most
instances the same schools have been sampled. Where
schools had to be replaced care was taken to find
equivalent schools to replace them. Evidence that this
strategy was successful comes from trends in height
and weight, which did not differ significantly between
IABIE V-Trend within agegroupper cohort in respiratonr conditionsfrom 197
over 13 vears
to 1986 andresultingestimateofpercentage itncrease inprevalence
I, Trenid in prcvalence
per anniiual cohort
0 Increase over 13 yscars
('959) confidernce intuval
",lI reind in prealencc
pei anniiual cohort
"O Iincrease over 13 sears
95916 cornfidence interval!
Asthma anid bronchitis
Persistent wvhcece, ISo asthma or hronchiiis
138-3 (90 8 to 197'4)
46-6 (9- 56 to
73-8 30 6to 131 1)
184 (31 4 to
378-4 (250-1 to554-2)
9522 (-62-6 to -38X8)
24-7 .14 to 48 9)
1167 (2295 to 146-4)
007 (-200 to23-3
76 3 2599to 146-9)
VOLUME 300 19 MAY 1990
the replaced areas and the areas that were in the study
throughout."4 This was not true in Scotland, but the
Scottish areas have for this reason not been included in
this analysis. Each school was visited as far as possible
in the same week each year. On rare occasions it was
necessary to change the date of the visit by up to a
maximum ofone month.
It is not surprising that the number of children with
a diagnosis of asthma has been increasing. Being
diagnosed as having asthma
receiving specific treatment for the condition,2 " and
there is already good reason to suppose that general
practitioners are seeingmore people who they recognise
as having asthma.' It could be argued, however, that
this represents an increasing awareness of asthma and
the importance of treating it rather than a genuine
increase in prevalence. Some support for this hypo-
thesis is given by the overall decrease in the prevalence
of asthma and bronchitis taken together. Such an
interpretation does not, however, explain the increase
in the prevalence of symptoms.
The greatest proportional increase among the boys,
and a substantial increase among the girls, was in those
with persistent symptoms but no diagnosis of either
asthma or bronchitis. In view of the close association
that is believed to exist between diagnostic labels and
treatment this is particularly worrving. It may indicate
an increasing number of untreated subjects with
prescription rates for the population as a whole. We do
not, however, have any information on the drugs that
were used over this period, and this remains only one
possible interpretation of the data.
The estimated change in the prevalence ofoccasional
wheeze is fairly small, and, though there is a positive
trend for each sex, only that for girls is significant
(p<0 05). The change in persistent wheeze is, however,
more substantial and is significant for boys and girls
(p<OOOl). There are four possible explanations for
this difference. Firstly, if a substantial proportion of
the population became slightly more prone to wheeze
this would show a greater increase in prevalence of
severe disease.' Secondly, wheeze is a symptom of a
heterogeneous group of conditions and children with
persistent wheeze "on most days and nights" may
represent a greater proportion of a particular subgroup
that has increased in size. For instance, more severe
wheeze and wheeze associated with atopic conditions
are more likely to persist through early childhood> and
adolescence'9 and be recognised as asthma.
possible that this category of wheeze, rather than
other more benign causes of childhood wheeze, has
increased. Thirdly, occasional wheeze is likely to be a
less specific marker of respiratory disorder than
persistent wheeze, and subjects may be more likely to
be misclassified when this marker is used than with the
more specific marker. This could reduce the sensitivity
of the question on wheeze for picking up changes over
time. Finally, we cannot exclude the possibility that
mothers have become more sensitive to their children's
svmptoms over this period and aresystematicallymore
likely to report wheeze as being "on most days or
nights." Burr et al showed that the prevalence ofsevere
grades of reactivity had increased more than that of
milder responses to exercise." The prevalence of a 15%/o
fall in peak expiratory flow rate after exercise increased
by 15% (from 6 7%/0 to 7-7%) over the 15 years between
their two surveys, but the prevalence of a 35% fall
in peak expiratory flow rate increased by 156%
(from 0 9%/, to 2 3%). This suggests that the greater
proportional increase in severe disease is not simply the
result of a reporting bias.
It is difficult to interpret direct comparisons between
in mortality and changes
Nevertheless, the increase in persistent wheeze over
is closely related to
this time is approximately 5% a year, a figure similar
to the increase noted in mortality from asthma in
5-34 year olds in the same period.' It might therefore be
possible to explain the whole of the increase in
mortality by the increase in prevalence, but the
intervals for all ofthese estimates should leave open the
question of whether changes in treatment may also
have affected mortality over this period.
in the prevalence of severe and
persistent wheeze is unexplained, but it may have been
due to a general increase in the prevalence of atopy.
Evidence for a concomitant increase in general practice
consultations with hay fever' and for a rapid increase in
the prevalence of eczema since 1946" support this
view. Such an increase in the prevalence ofatopy might
have been brought about by an increased exposure to
allergen, though there
environmental exposure that is known to alter this
vulnerability is tobacco smoke; in particular, smoking
in pregnant women has been associated with an
increased concentration of IgE in cord blood of
neonates and an increased incidence ofatopy in the first
year of life." Maternal smoking has also been shown to
be a risk factor for atopy later in adolescence," which
suggests that it leads to a long lived increase in atopy.
The increasing prevalence of smoking in women
of childbearing age up until the mid-1970s might
therefore explain some ofthe increase in atopic disease,
though there is insufficient information at present in
the national study of health and growth to test this
hypothesis adequately. An alternative hypothesis has
recently been advanced by Strachan, who showed a
strong negative association between the number of
older siblings and the prevalence of atopic disease in
children born in 1958." He suggested that lower
infection rates in early life may be associated with a
higher incidence ofatopy. There is little other evidence
to support this hypothesis at the moment, but if it is
correct the changes in family size over the last hundred
years could have led to a change in the incidence of
The increase in prevalence of morbidity reported in
this paper is sufficient to explain much of the increase
in mortality and use of services over this period. The
reasons for the increase are so far unknown, but large
changes in prevalence underline the importance of
understanding aetiology in order to formulate an
appropriate strategy for dealing with the current
is little evidence that this
administrators, nurses, and clerks in the areas and schools for
their participation and Professor W W Holland and the field
workers and administrators in our department for their work
in the study. The study is supported by the Department of
Health and the Scottish Home and Health Department.
Biriey P(iJ.Asthimia mortalityinEngland anid Wales: evidcncefor afurthcr
increase, 1974-84 Lanc.t 1986;ii:323-6.
2 Charlton JRH,
amnicablc to mcdiiccal intirvention.BrrMedj1986;292:295-3011
3 Andersoni HR, Bales' 1', West S 'I'rcnds in the hospital care of acute childhood
asthma 1970-8 a rcgional study. BrMIrM'd
Trends 198 7;49 1X-23
5 Fleming t)M, (rombic 1)L Prc\-alcnce of asthima and hay fever in Eiglatid
and Wales. Brided 71987;294:279-83.
6 Smith JM. '[lic prsaleinc of asthma and whccziig in childrcn. B1r7 DIts (Chest
7 \Vadssworth Intcrgcicrational diff'crnecs in child health.
Londtin: Otfic oil optilation Ccilstiscs and Stirs'cs, 1985
X Hill RA, Williamns J
scltool ahsetice in primars schooils. BrM.led7 1989:;299:898.
9 Btirr ML., Btitlandi 13K, King S, Vaughian-Willianis E. Chantges in asthilia
prcvalcince tuwo suirsies 15 ycars apart .4rch l)is Child 1989;64:1452-6.
113 HaY IFC, Higginbottamni
Lamn(ec 1987;ii 6119- 11
'4. [rinds ini morbiditv and mortalitv from astliTiia. Population
'I'attersficli Al, Brittmon JR. Asthma, sheezing, anti
Hals the inailagcnmlet of' asthimia improtvtec
19 MAY 1990
11 Anderson HR.
12 Hill RA, Stalndcn P'J, 'l'attersfield Al. Asthmiia, wheezing and school absclnce
in primary schools. Arch lis Child 1989;64:246-51
13 Dodge RR, Borrows B. The prevalence arid inicidence
asthma-likc symptoms in a general popuLlation sample. Am Reiz Rcspir )is
14 Jacksoii RT, Beaglchole R, Rca HH, Suthcrland l)C. Mortality from asthma: a
new epidetiic icii Ncw Zealand. BrMlcd] 1982;285:771-4.
15 lousqtiet J, Hattion F, (Jodard 1'. Michel IB. Asthma mortality in Fratice.
7 A// llrgv C/in Immunol 1987;80:389-94.
16 Evans R. Recetit adxances reflecting incrcases in mortality from asthma.
7A//creTsv Cliti Immunii/ 1987;80:377-9.
17 Mitchell EA, Ctletlr
)R. Paediatric admission to Atickland Hospital
asthma 1970-80. ,A/Mcdj 1984;97:67-70.
18 Halfon N, Newacheck PW. 'I'rends in the hospitalisation for actute childhood
asthma. Amz 7 Public hlealth 1986;76: 1308-1 1.
19 Chinn S, Rotia RJ. 'I'hc sectilar trend in the heiight of primars school children
in England and Scotland frrom 1972-80. Ann Hlum R3iol 1984;11: 1-16.
20 \Woolsont RF, Clarke WR. Analysis of categorical incompletc longitudilial
21 SAS Institute. SAS users' guide: statistics.
Carolina: SAS Itastittite, 1985:171-253.
22 Armitage 1', Bcrry G. Staotistical methods in mendital resetirch. 2nd cl. Oxf'ord:
Blackwell Scientific, 1987:386-99.
23 RonaRJ, AIttiman D)G. National study ot' hcaltlh and grovith: studics ot' attaitled
height, weight atid triceps skinfold in English children S to 11 years old.
Antn Hum BFiol 1977;4:501-23.
24 (Chinn S, Rona RJ, Price CE. Thc sccular tretid itt height of primarv school
Is the prevalence
olf asthma changing? Arch Dis (Child
of' asthmiia and
f ithe Roival Siaiiitica/lSociet)y 1984;147A:87-99.
crso)on S editiotn. Cars, Nortih
childreni in Eniglanid and Scotland 1972-1979 and 1979 to 1986. Ann liom
25 Speliglt ANP, Lce
I)A, Hey EN. Lnlderdiagniosis and undertrcatmncnt ot
asthmia in childlhood. BrAledj 19638286:1253-6.
26 Andcrsoni HR, Blaile
PA, Cooper JS, lI'altcr JC, Wcst S. Medical carc ol
asthma anid wheezing illlness in childrcn: a communit\ survec..7
C,onmunltvis Iclie.lh 1983 ;37 I1X0-6.
27 Mitchell EA. Is ctnrrenlt treatment increasing asthma mnortality and morbiditv?
28 Van Weel C, van den Bosch W'JHM, van deni Hoogen HLM,
Development of respirators illness in chilchoodii-
geineral practicc.JRCioll Gen Print 1987;37:404-8.
29 Giles GG, (uibsont HB,
'I'asmanian adolescents: a fotllow up (!f thc 1961 birth cohiort. ,4u .us/
30 Taylor B, Wadsworth J,
WVadsworth M, Peckham C. Changes in the
Lancet 1984;ii: 1255-7.
31 Magnttsson CG(. Maternal smoking influLences cord sertitt
and ittcreascs the risk for stibseqitent int'aiit allergy.
32 Ware JH, Diockery 1)W, Spiro EE, Ferris BGi. Passive smiking, gas cooking,
and respiratotry hcalth of childrcti livitng itt six cities. Aini Rev Reipir D)is
33 Strachan Dl). Hay fever, h\gienc, and hotischold size. Br Ved 7 1989;299:
a Ilingituldittal sitidv III
L-ickiss N, Shaw K. Respiratory svmptoirms
IgE and Igi) levels
Accepted 8 .Iarih 1990
Percutaneous cholecystolithotomy: the first 60 patients
S G Chiverton, J A Inglis, C Hudd, M J Kellett, R C G Russell, J E A Wickham
Objective-To assess the feasibility and possible
complications of percutaneous removal of gall
Design-Prospective study of the first 60 patients
Setting-The London Clinic.
Patients-60 Consecutive patients with sympto-
matic gall stones who agreed to have them removed
Results-56 Patients had stones
removed percutaneously. In four patients failure of
access necessitated a cholecystectomy under the
same anaesthetic. Two patients had an empyema of
the gall bladder drained initially, followed by a
second operation to remove the stones one week
later. Seven patients had postoperative complica-
tions, and two had recurrences of biliary calculi.
Conclusions-The techniques and instruments
used in percutaneous nephrolithotomy can success-
fully be adapted for percutaneous removal of gall
stones. The procedure is suitable for a wider range
of patients than other techniques that leave the gall
Percutaneous nephrolithotomy and extracorporeal
shock wave lithotripsv have now virtually replaced
open renal surgery. We decided to use the method
entailed in percutaneous nephrolithotomv to remove
calculi from the gall bladder. Other techniques are
available to clear biliarv calculi, but they have certain
disadvantages. Extracorporeal shock wave lithotripsy
and dissolution treatment both require the gall bladder
to be functioning and are restricted to particular types
of stones, and with both procedures the calculi take a
considerable time to clear.`4 Dissolution ofcalculi with
methyl tert-butyl ether requires percutaneous access
and is also time consuming.' Percutaneous cholecysto-
lithotomy allows stones to be completely removed
immediately and does not need a functioning gall
bladder. We have previously reported our initial
experience of the technique" and now report on our
first 60 patients.
Patients and methods
selected 29 men and 31 women aged 25-74 (mean age
51) for treatment. All had symptomatic gall stones.
Initially, we selected only patients whose gall bladder
was functioning, but later we included patients whose
gall bladders were not functioning if this was thought
to be due to a stone obstructing Hartmann's pouch.
None of the patients had any other medical conditions,
to have percutaneous cholecysto-
lithotomy after the alternatives had been explained.
Each patient agreed to have an immediate cholecystec-
tomy under the same anaesthetic ifthe procedure could
not be completed satisfactorily. We assessed the gall
oral cholecystography and ultrasono-
graphy, ultrasonography being used to determine its
accessibility and the line of approach needed for the
percutaneous track. Six hours before the operation
each patient was given oral contrast medium (calcium
ipodate 6 g) to aid fluoroscopic localisation of the gall
bladder. Prophylactic antibiotics were given when the
patients were anaesthetised.
1986 to September 1989 we
We anaesthetised patients on a fluoroscopic table
and placed drapes to collect the irrigant,
percutaneous nephrolithotomy. The gall bladder was
localised by both ultrasonography and fluoroscopy and
punctured, usually subcostally, with a 152 mm long
dwell sheathed needle (Becton Dickinson, Ontario,
Canada). A guidewire was introduced through the
sheath and a 7 Charriere gauge pigtail catheter (Cook
bladder to aspirate the bile and minimise any leakage.
The gall bladder was then filled with dilute contrast
medium and the tract dilated up to 28 Charriere gauge
with graduated metal dilators (Olympus, Keymed,
Southend on Sea) under fluoroscopic control. Finally,
an Amplatz sheath was inserted and the metal dilators
removed, leaving the guidewire in place. Care was
taken not to cause invagination of the gall bladder wall
during the dilatation.
We used an 18 French gauge nephroscope (Richard
of the gall bladder.
Small stones were removed
Department of Minimally
Invasive Surgery, London
Clinic, London WIN 2DH
S G Chiverton, FRCS, lecturer
J A Inglis, FRCS, lecturer
C Hudd, FRCS, lecturer
M J Kellett, FRCR, consultant
J E A Wickham, FRCS,
academic unit director
London WIP 7PN
R C G Russell, FRCS,
19 MAY 1990