Use of diagnosis-based risk adjustment models to predict individual health care expenditure under the National Health Insurance system in Taiwan.

BACKGROUND AND PURPOSE: Diagnostic information has been extensively studied and employed in the prediction of risk adjusted capitation payments in some countries. Nevertheless, few studies have been dedicated to the development of diagnosis-based risk adjusters in Taiwan. The purposes of this study were to develop outpatient diagnosis-based risk adjusters for a model of Taiwan's National Health Insurance (NHI) system and to evaluate the predictability of the risk adjustment models generated utilizing these adjusters. METHODS: Using a 2% random sample of 371,620 NHI enrollees, 5 risk adjustment models--i.e., demographic, inpatient diagnostic information outpatient diagnostic information, full diagnostic information, and prior utilization models--were evaluated with respect to predictive R2 and predictive ratios. While inpatient diagnosis-based risk adjusters were borrowed from previous research, outpatient diagnosis-based risk adjusters, referred to as Taiwan Ambulatory Spending Groups (TASGs), were developed based on 1996 claims data. RESULTS: The values of predictive R2 for the 5 risk adjustment models showed that the inclusion of outpatient diagnostic information considerably improved the predictability of the risk adjustment models for Taiwan's NHI system. Moreover, the predictive ratios revealed that the full diagnostic information model would reimburse different risk subgroups more fairly than the demographic, inpatient diagnostic information, and outpatient diagnostic information models and also outperform the prior utilization model with respect to disease risk groups. CONCLUSIONS: The risk adjustment model including the TASG risk adjusters can significantly improve predictability and can be employed to assess the NHI's current and proposed reform measures.     

Healthcare utilization patterns and risk adjustment under Taiwan's National Health Insurance system.

BACKGROUND AND PURPOSE: Some recent proposals for reform of Taiwan's National Health Insurance (NHI) system include risk adjustment mechanisms. However, there is a paucity of research on risk adjustment and its utilization in health insurance systems in Taiwan. The purposes of this study were to determine the healthcare utilization pattern and to develop a risk assessment model for capitation payments under NHI. METHODS: The individual enrollment and medical expenditure data for 1996 and 1997 were obtained from the Bureau of NHI. A random sample of 360,037 beneficiaries was divided into two sub-samples: one for model building and one for validation. Linear regression was employed to estimate the relationship between each individual's 1997 total expenditure and risk adjusters, i.e., age, gender, prior years' medical spending, and catastrophic status. RESULTS: The 10- to 14-years age group had the lowest total expenditure of $NT 3,055 ($US 1 = $NT 27.5 in 1996) in 1996, while the 65 years and over age group had the highest at almost 10 times more than the lowest. The distributions of total expenditure for both genders followed the familiar J-shaped curve. The average of the total expenditure of individuals with a catastrophic diagnosis was more than 17 times that of individuals without. Age and gender resulted in a predictive R2 of only 3.8% in the risk assessment model. By including prior total expenditure, the predictive R2 increased to 24.2%. Further addition of catastrophic status increased the predictability slightly to 25%. Prior outpatient expenditure predicted 72% of subsequent outpatient expenditure, but prior inpatient expenditure predicted only 3% of subsequent inpatient expenditure. CONCLUSIONS: As in other countries, age and gender provided only limited predictability in risk assessment. On the other hand, prior outpatient expenditure in this study provided relatively superior predictability in risk assessment. Prudence is required when including prior utilization as a part of the risk assessment model in calculating capitation payments, as this may indirectly encourage unnecessary use of healthcare services.     

Estimation of the financial burden to the National Health Insurance for patients with major cancers in Taiwan

BACKGROUND/PURPOSE: Almost all countries that have national health insurance schemes face financial challenges. A better understanding of the financial burden that cancer places on Taiwan's National Health Insurance (NHI) is important for helping policy makers to plan under scarce healthcare resources. This study attempts to estimate lifetime health expenditure for patients with 17 types of major cancers. METHODS: A total of 425,294 patients, each of whom was registered in Taiwan during 1990 to 2001 as having one of 17 major types of cancers, were included. All of them were followed until the end of 2004. Monte Carlo simulation was used to extrapolate survival for up to 600 months to derive the life expectancy or lifetime survival function after diagnosis for different cancers. The average annual health expenditure per case for each cancer type was calculated by using data from the NHI's reimbursement database. The lifetime health expenditure per case was estimated by multiplying the monthly survival probability by the average monthly health expenditure, adjusting for the annual discount rate and the medical care inflation rate. By incorporating the number of annual incidence cases, the total lifetime health expenditure can also be estimated. RESULTS: Of the 17 cancers studied, it was found that leukemia had the highest average annual health expenditure per case (207,000 TWD) as well as the highest lifetime health expenditure per case (2,404,000 TWD, without discounting adjustment). Breast cancer had the highest total lifetime health expenditure (5046 million TWD) because of the longer life expectancy and chronic morbidity. Furthermore, colorectal cancer had the second highest total lifetime health expenditure (4995 million TWD) due to its high incidence. CONCLUSION: The proposed method is a feasible way of estimating lifetime health expenditure for cancer patients even under high censoring rates. This would be helpful for cost-effectiveness assessment of cancer prevention programs and for policy planning.     

National health insurance expenditure for adult beneficiaries in Taiwan in their last year of life.

BACKGROUND AND PURPOSE: The cost of health care in the last year of life is a major issue of health services research. The purposes of this study were to examine health care use and expenditure under National Health Insurance (NHI) by Taiwanese adults in the last year of life, and to compare their results with those of randomly selected survivors. METHODS: A total of 9,369 adult decedent NHI beneficiaries were selected from NHI enrollee files and the Death Certificate data file for 1999. To compare expenses with those of patients without fatal illness, 10,000 randomly sampled adult beneficiaries of NHI who were alive on December 31, 1999, were selected as a survival group. NHI reimbursement and utilization information for these decedents and survivors were obtained by linking these samples to NHI claims files. RESULTS: The total NHI expenditure in the last year of life for the 9,369 decedents was US$71.6 million. About 54.5% of all medical expenses in the last year of life were incurred in the last 3 months of life. Nephritis and cancer were the most expensive causes of death, with per capita expenses of US$15,220 and US$10,828, respectively. The average expenses for survivors increased with age, while that for decedents increased with age from the 20- to 44-year age group to the 45- to 64-year age group, then decreased for decedents aged 65 and over. The age-dependent pattern of decreasing expenses with increasing age was found in cancer deaths for all age groups and in elderly decedent groups for most causes of death. CONCLUSIONS: This study demonstrated the relatively large amount of health care resources used by Taiwanese NHI participants in their last year of life and suggests the importance of greater awareness of the implications of allocation of medical care resources for terminally ill patients by policymakers and health care providers.     

The predictive value of a nonstress test taken 24 h before delivery in high-risk pregnancies

OBJECTIVES: A series of 180 cases of high risk pregnancies were studied in order to assess if a nonstress test taken 24 h before delivery is of any prognostic significance. METHODS: To assess the predictability of the NST (reactive or nonreactive) in terms of fetal outcome, the following variables were taken into consideration: fetal distress during labor, low Apgar score ( < 7) and perinatal mortality rate. RESULTS: A reactive test was found to be a good predictor of the healthy fetus (negative predictive value = 91.2%). Also, specificity of the test was found to be 85.4%. CONCLUSIONS: The nonreactive test could identify a population at risk but it was not helpful as a ‘stand alone’ modality in decision making, because of the low sensitivity and positive predictive value rates (40.9% and 28.1%, respectively).     

Predictability of pregnancy outcome in preterm delivery

Multivariate models have great potential value in enhancing the understanding of why some pregnancies have poor outcomes. Recently, such models have been advocated as a basis for predictive scoring systems that attempt to classify patients into high-risk and low-risk groups. In this report the usefulness of such an approach was assessed by studying the predictability of preterm delivery at The Johns Hopkins Hospital during 1980, using a multiple logistic model. Choosing a cutoff point (or probability of preterm delivery) of 10%, 697 of 2865 patients were placed in the high-risk group. The sensitivity, specificity, and positive predictive value of the model, as applied to this select population, were 62.2, 79.4, and 22.7%, respectively. Thus, only 23% of patients predicted to have preterm deliveries in fact delivered preterm. The predictive value could have been improved by increasing the cutoff point, but only at the expense of markedly reducing the sensitivity of the model. It was concluded that the potential value of multivariate analyses of pregnancy outcome as a predictive, risk-classification technique is limited. Nevertheless, such studies may aid the clinical evaluation of each individual patient by providing a better understanding of the etiologies of poor outcome.     

Crying without a cause and being easily upset in two-year-olds: heritability and predictive power of behavioral problems

In order to estimate the influence of genetic and environmental factors on 'crying without a cause' and 'being easily upset' in 2-year-old children, a large twin study was carried out. Prospective data were available for ~18,000 2-year-old twin pairs from the Netherlands Twin Register. A bivariate genetic analysis was performed using structural equation modeling in the Mx software package. The influence of maternal personality characteristics and demographic and lifestyle factors was tested to identify specific risk factors that may underlie the shared environment of twins. Furthermore, it was tested whether crying without a cause and being easily upset were predictive of later internalizing, externalizing and attention problems. Crying without a cause yielded a heritability estimate of 60% in boys and girls. For easily upset, the heritability was estimated at 43% in boys and 31% in girls. The variance explained by shared environment varied between 35% and 63%. The correlation between crying without a cause and easily upset (r = .36) was explained both by genetic and shared environmental factors. Birth cohort, gestational age, socioeconomic status, parental age, parental smoking behavior and alcohol use during pregnancy did not explain the shared environmental component. Neuroticism of the mother explained a small proportion of the additive genetic, but not of the shared environmental effects for easily upset. Crying without a cause and being easily upset at age 2 were predictive of internalizing, externalizing and attention problems at age 7, with effect sizes of .28-.42. A large influence of shared environmental factors on crying without a cause and easily upset was detected. Although these effects could be specific to these items, we could not explain them by personality characteristics of the mother or by demographic and lifestyle factors, and we recognize that these effects may reflect other maternal characteristics. A substantial influence of genetic factors was found for the two items, which are predictive of later behavioral problems.     

Early prediction of preterm birth for singleton, twin, and triplet pregnancies

OBJECTIVES: To create prediction models of early preterm birth for singletons, twin, and triplet pregnancies. STUDY DESIGN: We used a historical cohort study with the 1996 birth registration data for singletons and the 1995-1997 linked birth/infant death dataset for multiple births of the United States. Preterm birth was defined as gestational age <32 completed weeks. Eligible study subjects were randomly allocated to two groups: one group (80% subjects) for the creation of the prediction models, and the other group (20% subjects) for the validation of the established prediction models. Multivariate logistic regressions were used to establish the prediction models. We further assessed the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the established prediction models with different cut-off values in the validation group. RESULTS: The sensitivity, specificity, PPV, and NPV of the established model were 24.58, 93.54, 5.91, and 98.69%, respectively for singletons, 64.66, 57.04, 16.29, and 92.59%, respectively for twins, and 63.57, 53.58, 42.96, and 72.78%, respectively for triplets. CONCLUSION: The prediction models of early preterm birth for singleton, twin, and triplet pregnancies created by this study could be useful for obstetricians to identify women being at high risk of preterm birth at early gestation.     

Combined logistic and Bayesian modeling of cesarean section risk

OBJECTIVE: This study was undertaken to develop a simple and robust method for predicting risk of cesarean section. STUDY DESIGN: Retrospective cohort study of singleton births at term between 1992 and1999 in 22 Scottish maternity hospitals among primigravid women induced with prostaglandin. The risk of emergency cesarean section was modeled by using multivariate logistic regression in a development sample (n = 14,968). The output of this model was converted into adjusted likelihood ratios by using a novel method and tested in a validation sample (n = 12,638). RESULTS: Maternal age, height, gestational age, and fetal sex were all predictive of the risk of emergency cesarean section after prostaglandin induction of labor (all P < .001). The area under the receiver operator characteristic (ROC) curve in the development group was 0.677. The derived Bayesian model was comparably predictive of cesarean section risk in the validation sample: ROC of 0.673 (95% CI 0.662-0.684). Among the 994 women (8%) with a predicted cesarean section risk of more than 40%, the expected proportion was 48.2% and the observed proportion was 47.2%. Among the 1439 (11.4%) with a predicted cesarean risk of less than 10%, the expected proportion was 7.9% and the actual proportion 8.6%. CONCLUSION: Women at low or high risk of cesarean section after prostaglandin induction of labor can be identified with the use a novel combination of logistic regression and Bayesian modeling. The method is simple, robust, and may be generally applicable for clinical estimation of risk.     

Antenatal and intrapartum prediction of shoulder dystocia

Objectives

To (1) develop algorithms to calculate the risk of shoulder dystocia at individual deliveries; (2) evaluate screening for shoulder dystocia.

Study design

Retrospective analysis of 40284 consecutive term cephalic singleton pregnancies using a ‘train and test’ method. Four models were derived using logistic regression and tested (birthweight alone; birthweight and other independent antenatal variables; birthweight and all independent antenatal and intrapartum variables; and all independent variables excluding birthweight).

Results

Shoulder dystocia occurred in 240 deliveries (0.6%). Birthweight was the most important risk factor although 98 cases (41%) occurred in babies weighing <4.0 kg. Birthweight and maternal height were the only independent antenatal variables; for intrapartum use, only these and instrumental delivery were independent. The antenatal model could calculate an individual's risk; the intrapartum model could also calculate the risk if an instrumental delivery were undertaken. Both showed 0.7% women to have a risk of shoulder dystocia of >10%. Although the antenatal model had high predictability (area under curve 0.89), it was no better than birthweight alone and had a sensitivity of 52.4%. Where birthweight was excluded, prediction of shoulder dystocia was poor.

Conclusion

Antepartum and labour calculation of the risk of shoulder dystocia is possible. Whilst greatly hindered by the inaccuracy of estimating weight, it allows due weight to be given to factors which may already be influencing clinical practice. However, shoulder dystocia cannot be predicted with sufficient accuracy to allow universal screening.     

Risk factors for endometrial hyperplasia and cancer among women with abnormal bleeding

OBJECTIVE: To identify independent risk factors for endometrial neoplasia in women with abnormal perimenopausal or postmenopausal bleeding and to use those factors to develop and test a predictive model. METHODS: We conducted a case-control study of women with abnormal perimenopausal or postmenopausal bleeding who had endometrial samplings; cases had endometrial cancer or complex hyperplasia and controls had benign endometrial histologies. Multivariate logistic regression models identified factors associated with risks of endometrial neoplasia. The predictive abilities of our models and a published model were assessed using the area under receiver operating characteristic (ROC) curves, for which an area of 1.0 indicated perfect positive predictive ability and an area of 0.5 was expected by chance. RESULTS: There were 57 cases of endometrial hyperplasia or cancer and 137 controls. Parity was related inversely (odds ratio [OR] 0.70; 95% confidence interval [CI] 0.56, 0.88; P = .002) and weight directly (OR 1.02 per kg; 95% CI 1.01, 1.04; P = .018) to the risk of endometrial neoplasia. Age (OR 1.04 per year; 95% CI 1.00, 1.08; P = .06) and diabetes (OR 3.50; 95% CI 0.99, 12.33; P = .052) were significant marginally. The area under the ROC curve for our model was 0.75, indicating moderate predictive ability; the area under the ROC curve for the published model was lower at 0.66. CONCLUSION: Current clinical predictive models based on case-control studies do not have sufficient predictive ability to determine if women with abnormal perimenopausal or postmenopausal bleeding should have diagnostic testing.     

Risk adjustment for complications of hysterectomy: limitations of routinely collected administrative data.

OBJECTIVE: We sought to determine the utility of routinely collected administrative data for risk adjustment for complications of hysterectomy. STUDY DESIGN: Using abstracted discharge data on 107, 648 women undergoing hysterectomy in North Carolina from 1988 through 1994, we constructed logistic regression models for the prediction of medical and surgical complications incorporating coded demographic, diagnostic, and procedural data. RESULTS: The overall complication rate was 16%, with surgical complications (11.8%) more common than medical complications (6.7%). Hysterectomy type, teaching hospital status, patient age >/=65 years, and insurance status of Medicaid or no insurance were significantly associated with both medical and surgical complication risk, as were procedures performed for cancer or pregnancy complications. Models that incorporated coded comorbidity were better predictors of medical complications (C = 0.714) than surgical complications (C = 0.630). CONCLUSION: Although surgical complications of hysterectomy are more common than medical complications, risk adjustment methods that use routinely collected administrative data are better at predicting medical complications. Ambiguities in coding, misclassification, and uncoded factors such as disease severity limit the utility of administrative data for risk adjustment for hysterectomy complications.     

Exposure to chronic stress and ethnic differences in rates of bacterial vaginosis among pregnant women

OBJECTIVE: The purpose of this study was to assess the contribution of chronic social stressors to race/ethnic differences in the rate of bacterial vaginosis among pregnant women in the inner-city area. STUDY DESIGN: We conducted a cross-sectional clinical prevalence study in a sample of 2304 women at the first prenatal visit (14.8 +/- 0.2 weeks of gestation). Bacterial vaginosis was diagnosed by Nugent's method. Stress was measured at the individual and community levels with the use of interviews and administrative records. Logistic regression was used to assess the effects of stress on the odds of bacterial vaginosis occurrence, after adjustment for demographic and behavioral risk. RESULTS: Black women had significantly higher rates of bacterial vaginosis (64%) compared with white women (35%). Exposure to chronic stressors at the individual level differed by race (eg, 32% of the black women reported threats to personal safety compared with 13% of white women). There were significant racial differences in exposure to stress at the community level (eg, 63% of the black women lived in neighborhoods with aggravated assault rates that were above the citywide mean compared with 25% of the white women). After the adjustment for sociodemographic, behavioral risk, and perceived stress, the odds of the occurrence of bacterial vaginosis that was associated with the community level stressor of "homelessness" was significant (odds ratio, 6.7; 95% CI, 1.6-27.8). Inclusion of both individual and community level stressors reduced the black/white bacterial vaginosis odds ratio by 27%. CONCLUSION: Stressful exposures are associated positively with bacterial vaginosis in pregnancy in a sample of women of low income in the inner city. The measurement of stressors at multiple levels explained a significant proportion of the racial disparity in the rates of occurrence of bacterial vaginosis.     

Prediction of neonatal metabolic acidosis in women with a singleton term pregnancy in cephalic presentation: an external validation study

Objective?To externally validate two previously developed prognostic models that predict the risk for developing metabolic acidosis in newborns using both antepartum (model 1) and intrapartum (combined with antepartum, model 2) risk factors: parity, previous cesarean section, maternal diabetes mellitus, gestational age, induced onset of labor, meconium-stained amniotic fluid, and use of ST analysis.Study Design?The two prediction models were applied in women in active labor at more than 36 gestational weeks with singleton fetuses in cephalic presentation and with high-risk pregnancies (n?=?5049) who were included in a Swedish randomized trial between December 1, 1998, and June 4, 2000. The prognostic ability of the models was determined using calibration and discrimination measures.Results?Of 5049 infants in the validation population, 54 (1.1%) suffered from metabolic acidosis. After adjustment for incidence differences between the Dutch and Swedish cohorts, the prognostic models showed good calibration and moderate overall discrimination (C statistic 0.63, 95% confidence interval [CI] 0.55 to 0.71; and 0.64, 95% CI 0.55 to 0.72), for models 1 and 2, respectively).Conclusion?External validation of the clinical prediction models for metabolic acidosis in Swedish infants showed good calibration and moderate discriminative ability. Updating of the models to enhance their predictive abilities seems indicated.     

Prediction of low birth weight using modified Indian council of medical research antenatal scoring method

Abstract

Objective: To study the prediction of low birth weight (LBW) using modified Indian Council of Medical Research (ICMR) antenatal scoring method.

Method: The present longitudinal study was carried out amongst 1138 pregnant women residing in area covered by Kinaye primary health centre (PHC) in rural Karnataka, India.

Results: Modified ICMR risk scoring revealed that 597 (52.5%) women had a risk score 6–10 (mild risk), 142 (12.5%) women had risk score 11–15 (moderate risk) and 29 (2.5%) had risk score ≥16 (severe risk), whereas, remaining 370 (32.5%) had a score of 0–5 considered as “no risk group”. The incidence of LBW had direct relationship with the risk score. The sensitivity was high (80.6%), whereas, specificity was slightly low (70.4%), positive predictive value was low (43.8%) and negative predictive value high (92.7%) for LBW when the risk score cut-off point was >7.

Conclusion: The modified ICMR antenatal scoring method can be used at all levels of health care and is an ideal instrument for prediction of LBW at the community level. It can be easily applied by even a health worker, not time consuming and at the same time does not lose its predictability.     

Do microbiological factors account for poor pregnancy outcome among unmarried pregnant women in Poland?

OBJECTIVE: Being unmarried is a well-known risk factor for poor pregnancy outcome such as preterm delivery and intrauterine growth restriction. The aim of this prospective study was to assess the prevalence and risk of bacterial vaginosis (BV) and selected bacteria isolated from the lower genital tract and to determine the socioeconomic and microbiological characteristics that might be responsible for poor pregnancy outcome observed among unmarried pregnant women. METHODS: The study population comprised 196 pregnant women attending 10 randomly selected outpatient maternity units in the Lodz region, central Poland. Cervicovaginal samples were obtained between 8 and 16 weeks of gestation. Based on Spiegel's criteria, gram-stained vaginal smears were examined for BV and the BV-associated flora was sought by culture. To evaluate the risk factors, relative risk ratios were calculated using EPI INFO software. RESULTS: Among 196 pregnant women, 40 (20.4%) were unmarried. BV was diagnosed among 55 (28.1%) women studied.In the univariate analysis, unmarried pregnant women were characterized by younger age, primary educational level, poor economic situation and excessive smoking during pregnancy, as compared to married women. The unmarried status was a borderline risk factor for BV (OR = 1.83, 95% CI 0.94-4.9) after adjustment for age, smoking and education. An analysis of the microbiological culture from the lower genital tract revealed that unmarried pregnant women had a higher risk for several types of pathological microflora, as compared to married women. However, this finding was significant only for Mycoplasma hominis. The independent risk factors of M. hominis were the young age of the subject and a low concentration of Lactobacillus spp. CONCLUSIONS: The observed socioeconomic, demographic and microbiological differences between unmarried and married women could be responsible for the poor pregnancy outcome among unmarried pregnant women in Poland. Unmarried pregnant women should be covered by comprehensive medical care even before pregnancy.Further studies taking into account the role of psychological stress, patterns of sexual behavior and substance abuse during pregnancy could help identify the factors responsible for adverse pregnancy outcome among unmarried pregnant women.     

Predicting delivery within 48 hours in women treated with parenteral tocolysis

OBJECTIVE: To develop a prediction rule using clinical admission characteristics for women treated with parenteral tocolysis for preterm labor who are at highest risk of delivery within 48 hours. METHODS: We performed a case-control study of patients treated with magnesium sulfate for idiopathic preterm labor. A case was defined as a patient who received magnesium sulfate tocolysis and delivered within 48 hours of admission. We selected a 48-hour delay to delivery as a clinically relevant endpoint for the maximization of steroid benefit. Controls were patients who received magnesium sulfate tocolysis and remained undelivered 48 hours after admission. Cases and controls were identified by merging a pharmacy billing database with International Classification of Disease codes for premature labor. Medical records were reviewed and risk factor information was obtained. We focused on risk factors within the first hour of admission, because our goal was to identify patients at high risk of delivery early in their hospital course. Backward stepwise logistic regression was used to develop explanatory and predictive models. The focus of the predictive model was to maximize the test's sensitivity and negative predictive value. RESULTS: We identified 50 cases and 150 controls. The following six variables were included in the initial multivariable models based on bivariate analyses: white blood cell count at least 14.0 (1000/microL), cervical dilation at least 2 cm, bleeding, substance abuse, parity, and previous abortion. A two-variable model containing cervical dilation and bleeding had an overall accuracy of 73%, sensitivity of 62%, and specificity of 76%, and it was as sensitive and specific as more complex models. CONCLUSION: Cervical dilation of at least 2 cm and bleeding on admission had an overall accuracy of 73% in predicting the likelihood of delivery within 48 hours in women receiving magnesium sulfate.     

Anxiety and prevalence of psychiatric disorders among patients awaiting surgery for suspected ovarian cancer

Aim:  The goal of the current study was to determine the anxiety level and prevalence of psychiatric disorders among patients awaiting surgery for ovarian tumors. Also analyzed were the predictive factors for psychiatric disorders and changes after surgical diagnosis.
Methods:  Patients who underwent surgery for ovarian tumors were examined before and after surgery with the MINI International Neuropsychiatric Interview, the Spielberger State-Trait Anxiety Inventory (STAI) and the Mausley Personality Inventory (MPI). Participants diagnosed with cancer were examined a third time after being given an explanation about whether or not adjuvant chemotherapy was required.
Results:  Twenty-seven participants completed the study and were analyzed. Nine (33.3%) of these 27 participants were diagnosed as having adjustment disorder. There were no differences in the demographic data, STAI trait anxiety score and MPI score between the participants with or without adjustment disorder. At the pre-surgical interview, the STAI state anxiety score of the participants was high (49.5 ± 10.30). After pathological examination of the tumors, it was found that 12 patients had cancer (malignant group) and 15 patients had a benign tumor (benign group). At pre-surgery, the prevalence of adjustment disorder and the level of anxiety in the benign group were similar to those in the malignant group. There was a second surge of anxiety in patients who needed chemotherapy.
Conclusion:  The above findings demonstrate that patients with suspected ovarian cancer experience a high level of anxiety. Physicians should be aware of the risk of adjustment disorder in these patients. Additionally, ovarian cancer patients need psychological assessment during the course of treatment.     

Predicting cesarean delivery with decision tree models

OBJECTIVE: The purpose of this study was to determine whether decision tree-based methods can be used to predict cesarean delivery. STUDY DESIGN: This was a historical cohort study of women delivered of live-born singleton neonates in 1995 through 1997 (22,157). The frequency of cesarean delivery was 17%; 78 variables were used for analysis. Decision tree rule-based methods and logistic regression models were each applied to the same 50% of the sample to develop the predictive training models and these models were tested on the remaining 50%. RESULTS: Decision tree receiver operating characteristic curve areas were as follows: nulliparous, 0.82; parous, 0.93. Logistic receiver operating characteristic curve areas were as follows: nulliparous, 0.86; parous, 0.93. Decision tree methods and logistic regression methods used similar predictive variables; however, logistic methods required more variables and yielded less intelligible models. Among the 6 decision tree building methods tested, the strict minimum message length criterion yielded decision trees that were small yet accurate. Risk factor variables were identified in 676 nulliparous cesarean deliveries (69%) and 419 parous cesarean deliveries (47.6%). CONCLUSION: Decision tree models can be used to predict cesarean delivery. Models built with strict minimum message length decision trees have the following attributes: Their performance is comparable to that of logistic regression; they are small enough to be intelligible to physicians; they reveal causal dependencies among variables not detected by logistic regression; they can handle missing values more easily than can logistic methods; they predict cesarean deliveries that lack a categorized risk factor variable.     

Epidemiology and predictive values of risk factors for neonatal group B streptococcal sepsis

OBJECTIVES: To determine the incidence of and factors affecting risk factors for neonatal group B streptococcal (GBS) sepsis and their predictive values for intrapartum GBS carriage; to calculate the proportions of women eligible for intrapartum antibiotic prophylaxis (IAP) using different selection protocols. DESIGN: Cohort study. SETTING: Antenatal clinics and labour wards of a community hospital and a tertiary referral centre in western Sydney POPULATION: Women attending antenatal clinics during the study periods were invited to participate. METHODS: Approximately 500 women attending antenatal clinics were screened for GBS carriage at 26-32 weeks gestation and at delivery, using several screening methods. Clinical risk factors for neonatal sepsis were recorded during labour. MAIN OUTCOME MEASURES: Incidence of antenatal anovaginal GBS carriage and clinical risk factors during labour, their predictive values for intra-partum GBS carriage and their relationship, if any, to demographic and obstetric factors. RESULTS: Antenatal and intra-partum GBS carriage rates were similar but varied from 18% to 27%, depending on screening methods. The best positive and negative predictive values of antenatal GBS culture, for intra-partum carriage, were 69% (95% confidence interval (CI) 64-74) and 92% (95% CI 50-94) respectively Clinical risk factors occurred in similar proportions of GBS carriers and non-carriers. CONCLUSIONS: Neither early antenatal screening nor clinical risk factors are reliable predictors of intra-partum GBS carriage. Intra-partum antibiotic prophylaxis based on GBS carriage or risk factors when carrier status is unknown would involve approximately 35% of women, compared with approximately 16% if based on risk factors only Both strategies would prevent similar proportions of neonatal deaths from GBS sepsis. Compliance with a preventive protocol is the most likely determinant of its overall effectiveness.