British Journal of Obstetrics and Gynaecology December 1992, Vol. 99, pp. 1000-1003
CLINICAL PRACTICE
The usage of pregnancy tests in one health district S I M O N VOSS
Registrar in
Medicine Salisbury Health Authority, Salisbury Wiltshire SP2 8BJ, UK
ABSTRACT Objective To investigate the total number of pregnancy tests performed and the proportion that are repeated in one health district. To relate these results to the number of conceptions. Design Collection of data related to the numbers of pregnancy tests performed from all sources. Surveys of women booking for antenatal care or undergoing nonroutine obstetric ultrasound examination. Setting O n e health district in the south of England. Results About five pregnancy tests were performed for each proven pregnancy. Nearly two-thirds of tests were purchased over the counter, the remainder being carried out by hospital laboratories and pregnancy advisory services. Very few women had their pregnancy diagnosed on clinical grounds alone and a small number of ultrasound examinations were apparently performed in lieu of chemical tests. Repeated testing was more likely in primiparous women (P20 weeks).
Women undergoing ultrasound in early pregnancy During the 15 weeks of this study, 91 nonroutine ultrasound examinations in early pregnancy were performed. Most requests were from general practitioners (65%) and from the gynaecology ward for in-patients (29%). The remainder were from gynaecological, antenatal and family planning clinics. The indications for requests are shown in Table 3. A total of 17 (19%) women for whom an obstetric ultraTable 3. Indications for ultrasound request (n = 91).
Indication ? Pregnant ? Ectopic, including previous ectopic ? Spontaneous abortion ? cyst
IUCD ?Pregnant Back pain & PV bleeding
Other Total
Parity
Requests n
sound examination was requested had not had a laboratory pregnancy test. Of these, six requests were to exclude ectopic pregnancy, three to confirm pregnancy, three for possible spontaneous abortion (two of whom were not pregnant on ultrasound) and two for back pain and vaginal bleeding. The remainder were for different indications.
Discussion Pregnancy testing is a widely used service and is peculiar in that it is available from health care professionals, pharmacists and from supermarkets. The results from this study show that about seven pregnancy tests were performed for each live birth and about five for each proven pregnancy in the district in 1990. About 44 pregnancy tests were performed for every 100 women of childbearing age (15-44 years of age) during the year. If one assumes that the situation in the study district is similar to that elsewhere in England and Wales, the number of tests can be crudely extrapolated to 4.8 million pregnancy tests per annum in England and Wales. The cost of pregnancy testing kits used by the health authority was 12.5 pence per test for the Pregnosticon Planotest@ (low sensitivity haemagglutination) and 80 pence for the Hexagon@(high sensitivity ELISA) tests. The cost of performing a pregnancy test has been calculated using WelCan Units (Welsh Office 1989) which take account of staff costs, depreciation, heating, estate management, etc, and was f3.29 per test (average for both sorts of test, 3.45 WelCan units per test). The approximate cost to the health authority was therefore about El3 000 for the year of 1990.This study found that 17% of tests were repeats within a 4 month period. If one arbitrarily assumes that half of these may have been unnecessary, Ell00 could have been saved in this district if these tests had not been requested. The cost of pregnancy tests sold over the counter in the study district in 1990 was about f50 000, and therefore the largest proportion of the cost of pregnancy testing in the district was borne by the women themselves. The need for counselling with pregnancy testing, especially in adolescents, has been discussed previously (Stephenson 1989) and there have been suggestions that pharmacists should be doing more to promote the appropriate use of testing kits (Coons 1989). The increasing sale of the tests by supermarkets will make this more difficult. It is not known why so many pregnancy tests are per-
U S A G E O F P R E G N A N C Y TESTS
formed. Many women may be very keen to be pregnant, while others with amenorrhoea for whatever reason may be seeking reassurance that they are not. The fact that more primiparous women had more than one pregnancy test during the pregnancy may reflect their heightened apprehension. That so many pregnancy tests are performed may suggest that family planning advice is lacking or does not leave women confident that they are unlikely to have conceived. Repeated testing by health care professionals is easy to criticise but, faced with an anxious patient, a pregnancy test may be a good method of reassurance. A particular problem encountered during this study with self administered tests concerns their high sensitivity and the resulting ability to detect pregnancy earlier than routine hospital tests. This may produce the situation where a woman who has a positive home test does not trust this result and asks her doctor or the family planning clinic to confirm pregnancy. If it is still early in the pregnancy this second (lower sensitivity) test may return negative. This may confuse the woman, her doctor and the pharmacist. The doctor may then request a third test to finally confirm pregnancy. This is unsatisfactory and uneconomical for both the NHS and the woman. Only 2.8% of women who booked for maternity services during this survey had their pregnancy diagnosed clinically with no immunological test and some of these presented late in the pregnancy. The reasons why it is so important to have definite confirmation of pregnancy were discussed in a recent study (Pye &Thomas 1990).These include being able to plan accordingly, health-related reasons involving the adoption of a more healthy lifestyle and employment or family commitments. There is an obvious reason for early confirmation in women who wish to have the pregnancy aborted or where abnormal pregnancy is suspected. An American study has shown the low reliability of patient history in excluding pregnancy (Ramoska et al. 1989). There is little justification in providing pregnancy tests for women who present to a doctor with missed periods saying that they are trying to get pregnant and who have signs and symptoms of pregnancy. There seems to be a reluctance to accept the doctor’s clinical opinion and one wonders if the training of doctors has taught them to trust a laboratory test more than their own clinical judgement? In these circumstances perhaps a woman should purchase her own kit if she
1003
feels a need to confirm the pregnancy. The results of this study suggest, however, that even when a self test is performed many women often desire a hospital test which may be perceived as more accurate. The request for a pregnancy test in women who are clinically pregnant seems unnecessary and is an example of the medicalisation of society (Ilich 1975). Unless there are good medical reasons, women who want formal confirmation of pregnancy should be advised that the quickest way of achieving this is to purchase a self testing kit and they should be reassured that these tests, when performed correctly, are at least as sensitive as those performed in hospital.
Acknowledgments I gratefully acknowledge the help and advice of the Obstetric and Gynaecology Unit and the midwives, ultrasonographers, and pharmacists in Salisbury. I am also grateful to Dr N. Armand Smith, Dr S. Patrick, Stuart Mathews, Sarah Cook, Sister Day, Sister Powis, Jan Saunders and Dr Peter Old for their advice. References Coons S. J. (1989) A look at the purchase and use of home pregnancy-test kits. Am Pharm 29,4648. Edmonds D. K., et al. (1982) Early embryonic mortality in women. Fertil Steril38,447-453. Ilich I. (1975) Medical Nemesis: The Expropriation of Health. Calder and Boyars, London. OPCS (1991) The Public Health Common Data Set. HMSO, London. Pye M. & Thomas H. (1990) An assessment of a hospital based pregnancy testing service. J Pub H Med 12,186-189. Ramoska E. A., Sacchetti A. D. & Nepp M. (1989) Reliability of patient history in determining the possibility of pregnancyAnn Emerg Med 18,4&50. Salisbury Health Authority (1991) District clinical information. Stephenson J. N. (1989) Pregnancy testing and counselling. Pediatr Clin Am 36,681-696. Wilcox A. J., etal. (1988) Incidence of early loss of pregnancy. N Eng J Med 319,189-194. Received 28 April I992 Accepted 15 June 1992
CLINICAL PRACTICE
The usage of pregnancy tests in one health district S I M O N VOSS
Registrar in
Medicine Salisbury Health Authority, Salisbury Wiltshire SP2 8BJ, UK
ABSTRACT Objective To investigate the total number of pregnancy tests performed and the proportion that are repeated in one health district. To relate these results to the number of conceptions. Design Collection of data related to the numbers of pregnancy tests performed from all sources. Surveys of women booking for antenatal care or undergoing nonroutine obstetric ultrasound examination. Setting O n e health district in the south of England. Results About five pregnancy tests were performed for each proven pregnancy. Nearly two-thirds of tests were purchased over the counter, the remainder being carried out by hospital laboratories and pregnancy advisory services. Very few women had their pregnancy diagnosed on clinical grounds alone and a small number of ultrasound examinations were apparently performed in lieu of chemical tests. Repeated testing was more likely in primiparous women (P20 weeks).
Women undergoing ultrasound in early pregnancy During the 15 weeks of this study, 91 nonroutine ultrasound examinations in early pregnancy were performed. Most requests were from general practitioners (65%) and from the gynaecology ward for in-patients (29%). The remainder were from gynaecological, antenatal and family planning clinics. The indications for requests are shown in Table 3. A total of 17 (19%) women for whom an obstetric ultraTable 3. Indications for ultrasound request (n = 91).
Indication ? Pregnant ? Ectopic, including previous ectopic ? Spontaneous abortion ? cyst
IUCD ?Pregnant Back pain & PV bleeding
Other Total
Parity
Requests n
sound examination was requested had not had a laboratory pregnancy test. Of these, six requests were to exclude ectopic pregnancy, three to confirm pregnancy, three for possible spontaneous abortion (two of whom were not pregnant on ultrasound) and two for back pain and vaginal bleeding. The remainder were for different indications.
Discussion Pregnancy testing is a widely used service and is peculiar in that it is available from health care professionals, pharmacists and from supermarkets. The results from this study show that about seven pregnancy tests were performed for each live birth and about five for each proven pregnancy in the district in 1990. About 44 pregnancy tests were performed for every 100 women of childbearing age (15-44 years of age) during the year. If one assumes that the situation in the study district is similar to that elsewhere in England and Wales, the number of tests can be crudely extrapolated to 4.8 million pregnancy tests per annum in England and Wales. The cost of pregnancy testing kits used by the health authority was 12.5 pence per test for the Pregnosticon Planotest@ (low sensitivity haemagglutination) and 80 pence for the Hexagon@(high sensitivity ELISA) tests. The cost of performing a pregnancy test has been calculated using WelCan Units (Welsh Office 1989) which take account of staff costs, depreciation, heating, estate management, etc, and was f3.29 per test (average for both sorts of test, 3.45 WelCan units per test). The approximate cost to the health authority was therefore about El3 000 for the year of 1990.This study found that 17% of tests were repeats within a 4 month period. If one arbitrarily assumes that half of these may have been unnecessary, Ell00 could have been saved in this district if these tests had not been requested. The cost of pregnancy tests sold over the counter in the study district in 1990 was about f50 000, and therefore the largest proportion of the cost of pregnancy testing in the district was borne by the women themselves. The need for counselling with pregnancy testing, especially in adolescents, has been discussed previously (Stephenson 1989) and there have been suggestions that pharmacists should be doing more to promote the appropriate use of testing kits (Coons 1989). The increasing sale of the tests by supermarkets will make this more difficult. It is not known why so many pregnancy tests are per-
U S A G E O F P R E G N A N C Y TESTS
formed. Many women may be very keen to be pregnant, while others with amenorrhoea for whatever reason may be seeking reassurance that they are not. The fact that more primiparous women had more than one pregnancy test during the pregnancy may reflect their heightened apprehension. That so many pregnancy tests are performed may suggest that family planning advice is lacking or does not leave women confident that they are unlikely to have conceived. Repeated testing by health care professionals is easy to criticise but, faced with an anxious patient, a pregnancy test may be a good method of reassurance. A particular problem encountered during this study with self administered tests concerns their high sensitivity and the resulting ability to detect pregnancy earlier than routine hospital tests. This may produce the situation where a woman who has a positive home test does not trust this result and asks her doctor or the family planning clinic to confirm pregnancy. If it is still early in the pregnancy this second (lower sensitivity) test may return negative. This may confuse the woman, her doctor and the pharmacist. The doctor may then request a third test to finally confirm pregnancy. This is unsatisfactory and uneconomical for both the NHS and the woman. Only 2.8% of women who booked for maternity services during this survey had their pregnancy diagnosed clinically with no immunological test and some of these presented late in the pregnancy. The reasons why it is so important to have definite confirmation of pregnancy were discussed in a recent study (Pye &Thomas 1990).These include being able to plan accordingly, health-related reasons involving the adoption of a more healthy lifestyle and employment or family commitments. There is an obvious reason for early confirmation in women who wish to have the pregnancy aborted or where abnormal pregnancy is suspected. An American study has shown the low reliability of patient history in excluding pregnancy (Ramoska et al. 1989). There is little justification in providing pregnancy tests for women who present to a doctor with missed periods saying that they are trying to get pregnant and who have signs and symptoms of pregnancy. There seems to be a reluctance to accept the doctor’s clinical opinion and one wonders if the training of doctors has taught them to trust a laboratory test more than their own clinical judgement? In these circumstances perhaps a woman should purchase her own kit if she
1003
feels a need to confirm the pregnancy. The results of this study suggest, however, that even when a self test is performed many women often desire a hospital test which may be perceived as more accurate. The request for a pregnancy test in women who are clinically pregnant seems unnecessary and is an example of the medicalisation of society (Ilich 1975). Unless there are good medical reasons, women who want formal confirmation of pregnancy should be advised that the quickest way of achieving this is to purchase a self testing kit and they should be reassured that these tests, when performed correctly, are at least as sensitive as those performed in hospital.
Acknowledgments I gratefully acknowledge the help and advice of the Obstetric and Gynaecology Unit and the midwives, ultrasonographers, and pharmacists in Salisbury. I am also grateful to Dr N. Armand Smith, Dr S. Patrick, Stuart Mathews, Sarah Cook, Sister Day, Sister Powis, Jan Saunders and Dr Peter Old for their advice. References Coons S. J. (1989) A look at the purchase and use of home pregnancy-test kits. Am Pharm 29,4648. Edmonds D. K., et al. (1982) Early embryonic mortality in women. Fertil Steril38,447-453. Ilich I. (1975) Medical Nemesis: The Expropriation of Health. Calder and Boyars, London. OPCS (1991) The Public Health Common Data Set. HMSO, London. Pye M. & Thomas H. (1990) An assessment of a hospital based pregnancy testing service. J Pub H Med 12,186-189. Ramoska E. A., Sacchetti A. D. & Nepp M. (1989) Reliability of patient history in determining the possibility of pregnancyAnn Emerg Med 18,4&50. Salisbury Health Authority (1991) District clinical information. Stephenson J. N. (1989) Pregnancy testing and counselling. Pediatr Clin Am 36,681-696. Wilcox A. J., etal. (1988) Incidence of early loss of pregnancy. N Eng J Med 319,189-194. Received 28 April I992 Accepted 15 June 1992
