- Email: [email protected]
Validation of Prognostic Models
prospective study is necessary to better evaluate at which stage its prognostic value should be regarded as the best.
CLINICAL FEMALE INFERTILITY AND GYNECOLOGY Tuesday, October 24, 2000 P-274 Validation of Prognostic Models. P. F. Donderwinkel, F. Koenders, W. H. Pijffers, B. de Wolf, G. Kroon, M. J. Heineman. Division of Reproductive Medicine, Department of Obstetrics and Gynaecology, University Hospital Groningen, The Netherlands.
Results: 605 charts were available for review. 545 patients were caucasian (90%), 38 patients were black (6.2%) and 22 patients were placed in the “other” racial group. There were no significant differences in the mean age among the groups. Diagnoses are presented in Table 1. Blacks had significantly more tubal infertility (60% tubal disease and 40% previous tubal ligation) and significantly less ovulatory dysfunction than either of the other groups (p,.01). Patients without insurance had significantly more tubal disease and less ovulatory dysfunction (table 2). Of the black population, 40% did not have insurance while only 5% of the caucasian population were without insurance. Patients without insurance were 5.2 fold more likely to have had a tubal ligation.
Objectives: Prognostic Models (PM’s) are used to estimate the chance to conceive spontaneously. Before putting them into practice they need to be validated. The validation of PM’s in a cohort of couples with longstanding unexplained subfertility has been presented previously (1). Now we report the results of the validation of PM’s in a general cohort of subfertile couples. Design: Retrospective cohort study in a university-based hospital. Materials and Methods: In 1997 369 couples were registered at the outpatients clinic for Reproductive Medicine of the University Hospital Groningen. A routine subfertility investigation was performed. Exclusion criteria were: ovulatory disorders, azoospermia, bilateral tubal occlusion, any subfertility treatment and lost to follow-up. The chance to conceive spontaneously was calculated using PROGNOt (2) for the models Collins, Eimers, Comhaire and Wichman or by hand for the Snick-models. The estimated chance to conceive per PM was compared with the number of recorded pregnancies within one year after registration. Results: 238 couples fulfilled the inclusion criteria. The number of patients (n), the 95% confidence limits for the number of conceptions according to the estimated chance to conceive (EC) and the number of recorded pregnancies (RC) are presented. Prognostic chance group
0–10%
11–20%
21–30%
31–40%
.40%
Collins 102/2–11/8 101/7–21/15 19/1–9/8 6/0–4/2 2/0–2/0 Eimers 27/0–5/3 14/0–5/1 31/3–13/4 32/6–17/8 35/13–24/17 Comhaire 57/0–7/2 98/8–22/14 45/8–20/17 21/3–11/8 6/0–5/3 Wichmann 34/0–4/0 143/15–33/27 60/7–20/17 0 1/0–1/0 Snick I 68/0–7/5 48/2–11/3 43/6–18/6 0 52/15–28/17 Snick II 51/0–6/2 33/1–10/3 63/9–22/6 49/11–24/9 41/15–27/13*
Total 230/19–37/33 139/32–52/33 238/30–53/44 238/28–50/44 211/32–54/31* 237/49–74/33*
* Recorded number of conceptions is not within the 95% confidence limits of the predicted chance to conceive.
Conclusions: The models according to Snick predict more pregnancies than really occurred in this cohort. The PM’s according to Eimers, Collins, Comhaire and Wichman agree well with the recorded number of pregnancies. References: 1. Donderwinkel et al., Abstract P-700, Supplement Fertil Steril 1998, S347. 2. PROGNOt 1.0 bu MAHM Wiegerinck, The Parthenon Publishing Group, New York, 1996. P-275 Racial and Economic Demographics of Couples Seeking Infertility Treatment. J. C. Robins, J. Ach-Green, M. Scheiber, J. H. Liu, S. Awadalla, M. A. Thomas. Department of Obstetrics/Gynecology. University of Cincinnati Medical Center, Cincinnati, OH. Objectives: To examine the demographics of infertility patients by race and economic factors. Materials and Methods: A retrospective chart review was performed of all patients who presented for an infertility evaluation, between 1/99 and 12/99, to two infertility centers in Cincinnati, Ohio. Age, parity, race, diagnoses, and insurance status were determined. Patients were categorized into one of three groups by race (caucasian, black, other.) ANOVA was used to determine if population means were different and chi square was used to determine if the population proportions were different.
S182
Abstracts
Conclusions: Despite the fact that the city of Cincinnati has a black population of 18.3%, only 6.2% of the infertility patients were black. This is significantly less than expected, assuming that the incidence of infertility is uniformly distributed among races (p,.05). We believe this decrease is due to reduced access to health care, decreased awareness of available infertility services, and the perception that infertility services are unaffordable. Tubal infertility was more common in the black population and in patients without insurance coverage. This may be a reflection of lower socioeconomic status rather than race alone. These findings are concordant with NCHS data (Vital Health Stat 23(20) 1998) demonstrating that lower income women are more likely to have tubal ligation than those with higher income. It also may reflect the fact that health care workers provide inadequate contraceptive counseling to patients who may want to maintain future childbearing. P-276 Factors Influencing Pregnancy Rates With a Minimal Stimulation (MS) Protocol for Non-ART Infertility Treatment: Review of 413 Cycles. B. S. Houmard, P. Juang, V. Y. Fujimoto. Department of Obstetrics and Gynecology. University of Washington, Seattle, WA. Objectives: Minimal stimulation is a protocol first described for non-ART infertility treatment in 1996 (Fertil Steril, 1996;65:583–587), consisting of 5 days of clomiphene citrate (100 mg/day) followed by a single dose of gonadotropin (150 U) on cycle day 9. In our clinic, this protocol has been used in combination with intrauterine insemination (IUI). The published description of this protocol (61 women, 106 cycles) showed a favorable pregnancy rate (20.8%) in a young population (mean age 5 31.9) with a high rate of ovulatory dysfunction (40%). The objectives of the current study were: (1) to determine the effectiveness of this protocol in the general infertility population cared for in a university-based infertility practice and (2) to evaluate factors influencing pregnancy rates obtained with this MS protocol. Design: A retrospective chart review of all individuals undergoing MS from 1997 through 1999 at our institution was conducted. Treatment cycles with documentation of a first trimester pregnancy outcome were included in the analyzes.
Vol. 74, No. 3, Suppl. 1, September 2000
CLINICAL FEMALE INFERTILITY AND GYNECOLOGY Tuesday, October 24, 2000 P-274 Validation of Prognostic Models. P. F. Donderwinkel, F. Koenders, W. H. Pijffers, B. de Wolf, G. Kroon, M. J. Heineman. Division of Reproductive Medicine, Department of Obstetrics and Gynaecology, University Hospital Groningen, The Netherlands.
Results: 605 charts were available for review. 545 patients were caucasian (90%), 38 patients were black (6.2%) and 22 patients were placed in the “other” racial group. There were no significant differences in the mean age among the groups. Diagnoses are presented in Table 1. Blacks had significantly more tubal infertility (60% tubal disease and 40% previous tubal ligation) and significantly less ovulatory dysfunction than either of the other groups (p,.01). Patients without insurance had significantly more tubal disease and less ovulatory dysfunction (table 2). Of the black population, 40% did not have insurance while only 5% of the caucasian population were without insurance. Patients without insurance were 5.2 fold more likely to have had a tubal ligation.
Objectives: Prognostic Models (PM’s) are used to estimate the chance to conceive spontaneously. Before putting them into practice they need to be validated. The validation of PM’s in a cohort of couples with longstanding unexplained subfertility has been presented previously (1). Now we report the results of the validation of PM’s in a general cohort of subfertile couples. Design: Retrospective cohort study in a university-based hospital. Materials and Methods: In 1997 369 couples were registered at the outpatients clinic for Reproductive Medicine of the University Hospital Groningen. A routine subfertility investigation was performed. Exclusion criteria were: ovulatory disorders, azoospermia, bilateral tubal occlusion, any subfertility treatment and lost to follow-up. The chance to conceive spontaneously was calculated using PROGNOt (2) for the models Collins, Eimers, Comhaire and Wichman or by hand for the Snick-models. The estimated chance to conceive per PM was compared with the number of recorded pregnancies within one year after registration. Results: 238 couples fulfilled the inclusion criteria. The number of patients (n), the 95% confidence limits for the number of conceptions according to the estimated chance to conceive (EC) and the number of recorded pregnancies (RC) are presented. Prognostic chance group
0–10%
11–20%
21–30%
31–40%
.40%
Collins 102/2–11/8 101/7–21/15 19/1–9/8 6/0–4/2 2/0–2/0 Eimers 27/0–5/3 14/0–5/1 31/3–13/4 32/6–17/8 35/13–24/17 Comhaire 57/0–7/2 98/8–22/14 45/8–20/17 21/3–11/8 6/0–5/3 Wichmann 34/0–4/0 143/15–33/27 60/7–20/17 0 1/0–1/0 Snick I 68/0–7/5 48/2–11/3 43/6–18/6 0 52/15–28/17 Snick II 51/0–6/2 33/1–10/3 63/9–22/6 49/11–24/9 41/15–27/13*
Total 230/19–37/33 139/32–52/33 238/30–53/44 238/28–50/44 211/32–54/31* 237/49–74/33*
* Recorded number of conceptions is not within the 95% confidence limits of the predicted chance to conceive.
Conclusions: The models according to Snick predict more pregnancies than really occurred in this cohort. The PM’s according to Eimers, Collins, Comhaire and Wichman agree well with the recorded number of pregnancies. References: 1. Donderwinkel et al., Abstract P-700, Supplement Fertil Steril 1998, S347. 2. PROGNOt 1.0 bu MAHM Wiegerinck, The Parthenon Publishing Group, New York, 1996. P-275 Racial and Economic Demographics of Couples Seeking Infertility Treatment. J. C. Robins, J. Ach-Green, M. Scheiber, J. H. Liu, S. Awadalla, M. A. Thomas. Department of Obstetrics/Gynecology. University of Cincinnati Medical Center, Cincinnati, OH. Objectives: To examine the demographics of infertility patients by race and economic factors. Materials and Methods: A retrospective chart review was performed of all patients who presented for an infertility evaluation, between 1/99 and 12/99, to two infertility centers in Cincinnati, Ohio. Age, parity, race, diagnoses, and insurance status were determined. Patients were categorized into one of three groups by race (caucasian, black, other.) ANOVA was used to determine if population means were different and chi square was used to determine if the population proportions were different.
S182
Abstracts
Conclusions: Despite the fact that the city of Cincinnati has a black population of 18.3%, only 6.2% of the infertility patients were black. This is significantly less than expected, assuming that the incidence of infertility is uniformly distributed among races (p,.05). We believe this decrease is due to reduced access to health care, decreased awareness of available infertility services, and the perception that infertility services are unaffordable. Tubal infertility was more common in the black population and in patients without insurance coverage. This may be a reflection of lower socioeconomic status rather than race alone. These findings are concordant with NCHS data (Vital Health Stat 23(20) 1998) demonstrating that lower income women are more likely to have tubal ligation than those with higher income. It also may reflect the fact that health care workers provide inadequate contraceptive counseling to patients who may want to maintain future childbearing. P-276 Factors Influencing Pregnancy Rates With a Minimal Stimulation (MS) Protocol for Non-ART Infertility Treatment: Review of 413 Cycles. B. S. Houmard, P. Juang, V. Y. Fujimoto. Department of Obstetrics and Gynecology. University of Washington, Seattle, WA. Objectives: Minimal stimulation is a protocol first described for non-ART infertility treatment in 1996 (Fertil Steril, 1996;65:583–587), consisting of 5 days of clomiphene citrate (100 mg/day) followed by a single dose of gonadotropin (150 U) on cycle day 9. In our clinic, this protocol has been used in combination with intrauterine insemination (IUI). The published description of this protocol (61 women, 106 cycles) showed a favorable pregnancy rate (20.8%) in a young population (mean age 5 31.9) with a high rate of ovulatory dysfunction (40%). The objectives of the current study were: (1) to determine the effectiveness of this protocol in the general infertility population cared for in a university-based infertility practice and (2) to evaluate factors influencing pregnancy rates obtained with this MS protocol. Design: A retrospective chart review of all individuals undergoing MS from 1997 through 1999 at our institution was conducted. Treatment cycles with documentation of a first trimester pregnancy outcome were included in the analyzes.
Vol. 74, No. 3, Suppl. 1, September 2000