[Implementing new interventions and indications: the possible role for real world data]. / De mogelijke rol van 'real world data'.

In advising the preferred therapy for the individual patient the expected results of the proposed intervention and possible side effects are the most relevant considerations. However, predicting the results of an intervention is difficult, especially when well designed randomized clinical trials (RCT's) are lacking or not conclusive. Artificial intelligence (AI) algorithms based on routine clinical data (real world data) can support clinical decision making, but in daily practice AI is still scarcely used. In this article one large radiotherapy facility and two health insurers describe their joint opinion on the possible role of AI based on real world data as an aid in clinical decision making when evidence from RCT's is not available. The introduction of proton radiotherapy in The Netherlands is being used as case model for AI model based clinical decision making.

Sinonasal cancer in The Netherlands: Follow-up of a population-based study 1989-2014 and incidence of occupation-related adenocarcinoma.

BACKGROUND: Long-term trends of sinonasal cancer in The Netherlands have been investigated with particular attention on adenocarcinoma for which wood or leather dust is a well-known risk factor. METHODS: All 4345 patients (1989-2014) registered in the Netherlands Cancer Registry were included. Standardized 3-year moving incidence rates per 1 000 000/person-years, and estimated annual percentage change (EAPC) were calculated. RESULTS: Forty-seven percent of the patients had squamous cell carcinoma (SCC), 12% had lymphoma, and 12% had adenocarcinoma. Sixty-one percent of the tumors were located in the nasal cavity, 22% in the maxillary, and 11% in the ethmoidal sinus. Male incidence decreased to 11.5/1 000 000 due to less SCC (EAPC -0.9%; 95% confidence interval [CI] -1.6 to 0.3) and adenocarcinoma (EAPC -4.3%; 95% CI -5.5 to 3.1). Female incidence increased to 7/1 000 000 (EAPC +2.0%; 95% CI +1.1 to +3.0) due to more SCC (EAPC +2.2%; 95% CI +1.0 to +3.5), whereas adenocarcinoma remained stable (0.6/1 000 000; EAPC +1.1%; 95% CI -6.0 to +8.7). Tumors of the nasal cavity increased in women (EAPC +3.3%; 95% CI +2.0 to 4.7). CONCLUSION: The decrease of male sinonasal adenocarcinoma may be the result of preventive measures combined with less workers in high-risk occupations.

Employment and insurance outcomes and factors associated with employment among long-term thyroid cancer survivors: a population-based study from the PROFILES registry.

PURPOSE: To obtain insight into employment and insurance outcomes of thyroid cancer survivors and to examine the association between not having employment and other factors including quality of life. METHODS: In this cross-sectional population-based study, long-term thyroid cancer survivors from the Netherlands participated. Clinical data were collected from the cancer registry. Information on employment, insurance, socio-demographic characteristics, long-term side effects, and quality of life was collected with questionnaires. RESULTS: Of the 223 cancer survivors (response rate 87 %), 71 % were employed. Of the cancer survivors who tried to obtain insurance, 6 % reported problems with obtaining health care insurance, 62 % with life insurance, and 16 % with a mortgage. In a multivariate logistic regression analysis, higher age (OR 1.07, CI 1.02-1.11), higher level of fatigue (OR 1.07, CI 1.01-1.14), and lower educational level (OR 3.22, CI 1.46-7.09) were associated with not having employment. Employment was associated with higher quality of life. CONCLUSIONS: Many thyroid cancer survivors face problems when obtaining a life insurance, and older, fatigued, and lower educated thyroid cancer survivors may be at risk for not having employment.

Trends in sinonasal cancer in The Netherlands: more squamous cell cancer, less adenocarcinoma. A population-based study 1973-2009.

BACKGROUND: Cancer of the nasal cavity or the paranasal sinuses (sinonasal cancer) is rare. Sinonasal cancer has been associated with various occupational risk factors such as exposure to dust of hard wood and leather. Also, a relationship with smoking habits has been suggested. We studied the long term trends in incidence to evaluate a putative effect of past preventive measures or changes in risk factors. DESIGN: A retrospective population-based descriptive study. OBJECTIVE: To interpret the long term trends in incidence of sinonasal cancer in The Netherlands. METHODS: Data of all 3329 patients >15 years registered during 1989-2009 by the Netherlands Cancer Registry (NCR) were analysed, by data of 447 patients registered by the Eindhoven Cancer Registry (ECR) during 1973-2009 were analysed separately. Information on patients and tumour characteristics was obtained from both registries. The incidence was calculated per 1,000,000 person years and standardised using the European Standard Population. RESULTS: Squamous cell carcinoma (SCC) was the most prominent histological type (48%), followed by adenocarcinoma (15%) and melanoma (8%). SCC was more frequently located in the nasal cavity or sinus maxillaris, but adenocarcinoma was more located in the ethmoid sinus. The male incidence increased during 1973-1995 with a peak of 15/1,000,000/year, decreasing since then to 11/1,000,000/year due to a declining incidence of both SCC and adenocarcinoma. In females the incidence remained stable around 5/1,000,000/year up to 2006 and increased to 7.5/1,000,000 in 2009 as a result of more SCC. The male/female ratio for SCC decreased from 2.7 to 2.0, and for adenocarcinoma from 3.4 to 2.8 since 1989. CONCLUSIONS: The higher incidence in males and the different trends in incidence in males and females may reflect differences in previous exposure to risk factors. Adenocarcinoma, related to occupational exposures, tend to decline. The trends in both male and female sinonasal SCC are comparable with the trends in lung cancer.

[Surgery and follow-up 131I therapy in patients with differentiated thyroid carcinoma in the south-east of the Netherlands, 1983-1996, compliance with the 1987 consensus guidelines: additional data]. / Chirurgie en 131I-nabehandeling bij patiënten met gedifferentieerde schildklierkanker in Zuidoost-Nederland, 1983-1996, vergeleken met de consensusrichtlijnen uit 1987: aanvullende gegevens.

In a previous study using data from the regional cancer registry of the Comprehensive Cancer Centre South, Eindhoven, The Netherlands, we concluded that in the majority of cases surgical treatment was in accordance with the consensus recommendations, but that about 40% of patients with differentiated thyroid cancer from a number of regional hospitals had not been referred for 131I therapy. However, in a subsequent study using patient data from these hospitals, it became clear that almost all patients had in fact been referred for therapy but to centres outside the 131I therapy region. The conclusion of the study should therefore be altered: the great majority of patients with differentiated thyroid cancer in the south-east of The Netherlands (1983-96) were referred for 131I treatment and therefore the primary surgical and the follow-up treatment complied with the 1987 consensus guidelines.

Feasibility of a randomized trial on adjuvant radio-iodine therapy in differentiated thyroid cancer.

BACKGROUND: Justification for adjuvant radio-iodine (I-131) therapy in differentiated thyroid cancer (DTC) is purely based on retrospective data. This is true for ablative therapy and even more so for high-dosage adjuvant schedules. Randomized trials on the latter application are considered impossible due to anticipated formidable sample sizes required in a disease with an overall excellent prognosis like DTC. OBJECTIVE: To develop and validate a model that could stratify for risk of recurrence, rather than survival, as is usually done in prognostic indices, and secondly, to use this model to estimate the sample size required for a randomized trial. DESIGN, PATIENTS AND RESULTS: From databases of three large Dutch centres, we identified 342 consecutive patients without known residual DTC after (near-) total thyroidectomy. Using Cox proportional hazards analysis, a model was validated that clearly distinguished risk categories of recurrence using commonly available baseline variables. The model included age, N stage at presentation and T stage in papillary carcinoma. According to this stratification, a subset of patients at substantial risk for relapse (30-40%) was identified. They could be eligible for a trial assessing the impact of high-dose adjuvant I-131 on recurrence rates. Assuming a clinically relevant effect of 30% reduction of relapses, 290 patients would have to be entered in either arm (alpha 0.05, power 80%). CONCLUSION: We conclude that even though a randomized trial on this issue will be difficult to design and conduct, sample size is not the main problem.

[Surgery and referral for subsequent 131I therapy for patients with differentiated thyroid carcinoma in the south-east of the Netherlands, 1983-1996, compared to the consensus guidelines from 1987]. / Chirurgie en 131I-nabehandeling bij patiënten met gedifferentieerde schildklierkanker in Zuidoost-Nederland, 1983-1996, vergeleken met de consensusrichtlijnen uit 1987.

OBJECTIVE: To evaluate the treatment of patients with differentiated (papillary or follicular) thyroid cancer in general hospitals in the south-east of the Netherlands during the period 1983-1996, in relation to the 1987 national consensus recommendations. DESIGN: Population-based, retrospective, descriptive. METHOD: For the period 1 January 1983-31 December 1996, data on the histology, TNM-stage and treatment (hospital, specialist, type of operation, referral for 131I therapy) of all 236 patients with differentiated thyroid cancer were obtained from the cancer registry of the Comprehensive Cancer Centre South, Eindhoven, the Netherlands. The treatment was compared with the recommendations from the consensus meeting in 1987. RESULTS: Data on 219 patients (137 papillary, 82 follicular thyroid carcinoma) treated in the general hospitals in the region were studied; the 17 remaining patients had been referred from outside the region. Patients were treated at all hospitals in the region; the number of specialists per hospital able to treat thyroid carcinoma (internist and/or surgeon) was limited. In total 79% of the patients underwent a (near-)total thyroidectomy, half of them in two phases, and in 12% of the cases combined with regional lymph node dissection. In the majority of cases, surgical treatment was in accordance with the consensus recommendations: 65-100% of the cases per hospital. The proportion of patients referred for 131I therapy varied from 17% to 90%; referral was more frequent in the case of larger tumours and/or metastases. Of the 24 patients with a small papillary carcinoma without metastases, 79% were not referred for 131I therapy. CONCLUSIONS: The recommendations laid down in the consensus meeting in 1987 were known and appeared to be followed for surgical treatment but for subsequent 131I therapy they appeared to be interpreted differently. A review of the consensus guidelines seems to be worthwhile.

Thyroid peroxidase antibodies during gestation are a marker for subsequent depression postpartum.

OBJECTIVE: Depression is not adequately diagnosed in many cases. Therefore, the question arises as to whether markers exist for depression. We investigated whether the presence of thyroperoxidase antibodies (TPOAbs) during pregnancy can be regarded as a marker for depression in the first year postpartum, particularly in relation to (overt or subclinical) thyroid dysfunction and other determinants of depression. DESIGN: This work was a prospective observational study. PATIENTS: A cohort of 310 unselected women (residing in the Kempen Region, southeastern Netherlands) were visited at 12 and 32 weeks gestation and at 4, 12, 20, 28 and 36 weeks postpartum. METHODS: At each visit, TSH, free thyroxine and TPOAb testing was performed, determinants associated with depression were asked for, and depression was assessed (according to the Research Diagnostic Criteria). Multiple logistic regression was performed to determine independent risk factors (odds ratios, ORs) for depression in gestation and/or postpartum depression. RESULTS: Data for 291 women were available for analysis; 41 women (14.1%) had TPOAbs at one or more time points, and 117 women (40.1%) had depression at one or more time points postpartum. The multiple logistic regression analysis showed that TPOAbs were independently associated with depression at 12 weeks gestation and at 4 and 12 weeks postpartum (OR, 95% confidence interval: 2.4 (1.1-6.0), 3.8 (1.3-7.3) and 3.6 (1.2-7.1) respectively). After the exclusion of women who were depressed at 12 weeks gestation (n=70), the presence of TPOAbs during early pregnancy was still found to be associated with the development of postpartum depression (OR, 95% confidence interval: 2.8 (1.7-4.5); after exclusion of women who had had depression in earlier life (n=51), TPOAb during early gestation was still associated with postpartum depression (OR, 95% confidence interval: 2.9 (1.8-4.3). CONCLUSIONS: The presence of TPOAbs during gestation is associated with the occurrence of subsequent depression during the postpartum period and as such can be regarded as a marker for depression.

Low maternal free thyroxine concentrations during early pregnancy are associated with impaired psychomotor development in infancy.

BACKGROUND: Maternal thyroid function during early pregnancy is an important determinant of early fetal brain development because the fetal thyroid is unable to produce any T4 before 12-14 weeks' gestation. Overt maternal hypothyroidism as seen in severe iodine-deficient areas is associated with severely impaired neurological development of the offspring. At present, it is not known whether low free T4 (fT4) levels during pregnancy in healthy women from iodine sufficient areas may affect fetal neurodevelopment. METHODS: Neurodevelopment was assessed at 10 months of age in a cohort of 220 healthy children, born after uncomplicated pregnancies and deliveries, using the Bayley Scales of Infant Development. Maternal TSH, fT4 and TPO antibody status were assessed at 12 and 32 weeks' gestation. Maternal gestational fT4 concentration was defined as an independent parameter for child development. RESULTS: Children of women with fT4 levels below the 5th (< 9.8 pmol/l, n = 11) and 10th (< 10.4 pmol/l, n = 22) percentiles at 12 weeks' gestation had significantly lower scores on the Bayley Psychomotor Developmental Index (PDI) scale at 10 months of age, compared to children of mothers with higher fT4 values (t test, mean difference: 14.1, 95% confidence interval (CI): 5.9-22 and 7.4, 95% CI: 1.1-13.9, respectively). At 32 weeks' gestation, no significant differences were found. In the group of women with the lowest 10th percentile fT4 concentrations at 12 weeks' gestation, a positive correlation was found between the mothers' fT4 concentration and children's PDI scores (linear regression, R: 0.46, P = 0.03). After correction for confounding variables, a fT4 concentration below the 10th percentile at 12 weeks' gestation was a significant risk factor for impaired psychomotor development (RR): 5.8, 95% CI: 1.3-12.6). CONCLUSIONS: Low maternal plasma fT4 concentrations during early pregnancy may be an important risk factor for impaired infant development.

Trends in treatment and long-term survival of thyroid cancer in southeastern Netherlands, 1960-1992.

Thyroid cancer (TC), comprising less than 1% of all cancers in the Netherlands, has a good prognosis in general. Controversy still remains on the extent of surgical treatment and the indication for additional Iodine-131 (131I) therapy in the management of differentiated TC. The aim of this study was to describe (changes in) the treatment of TC and to determine independent prognostic factors for crude and relative survival of differentiated TC diagnosed in general hospitals. This population-based, retrospective study was based on data from the Eindhoven Cancer Registry, Comprehensive Cancer Centre South (I.K.Z.), Eindhoven, the Netherlands. Data were collected on all 343 TC patients diagnosed from 1 January 1960 to 31 December 1992. All available information on treatment (initial and additional) and survival (on 1 April 1994) were recorded. Initial surgical treatment was defined as limited or extended. Multivariate analysis of crude and relative survival to determine prognostic factors for differentiated TC was performed. Mean follow-up was 7.6 years. The proportion of patients with differentiated TC increased from 60% in 1960-1972 to 84% in 1985-1992. TC patients were treated in all hospitals in the region, approximately 2-4/year. Ninety per cent of all TC patients initially underwent surgical treatment; the extended procedures increasing from 27% in 1960-1974 to 61% in 1985-1992. 131I was also administered increasingly (from 18-44%) to patients with differentiated TC. The relative 5, 10 and 20 year survival rates for all TC were 80, 75 and 75%, respectively. In the first 5 years after diagnosis the crude death ratio was higher with the rise of age and for the follicular type and after 5 years for males and advanced disease. After inclusion of surgical treatment into the model, the estimates of the other death ratios did not change. Patients treated with 131I did better only during the first 5 years. Although the prognosis for TC patients treated in general hospitals in Southeastern Netherlands was similar to that found for patients treated in referral centres, concentration of treatment should be considered.

Prediction of post partum thyroid dysfunction: can it be improved?

BACKGROUND: Screening pregnant women for thyroid peroxidase antibodies (TPOAb) to identify those at risk for post partum thyroid dysfunction (PPTD) is controversial, mainly because of the low positive predictive value (ppv) of TPOAb. OBJECTIVES: To evaluate if the ppv of TPOAb can be enhanced, either by taking into account the time of TPOAb testing, or by combining this parameter with other putative determinants of PPTD such as smoking, family history or other autoimmune diseases. METHODS: A prospective study was performed in the Kempenland region (southeastern Netherlands). Three hundred and ten unselected women were visited at 12 and 32 weeks gestation and 4, 12, 20, 28 and 36 weeks post partum. Serial thyroid stimulating hormone (TSH), free thyroxine (fT4) and TPOAb testing was performed. Thyroid dysfunction (TD) was defined as abnormal TSH either in combination with abnormal fT4 (overt TD) or without abnormal fT4 (subclinical TD). PPTD was defined as overt TD post partum. Multivariate regression analysis was performed for determining independent risk factors for PPTD. The sensitivity and specificity of TPOAb at different time points and at different concentrations were calculated and presented in receiver operating characteristic (ROC) curves. Women who had experienced PPTD were followed for 2.5-3 years. RESULTS: Data from 291 women were available for analysis. Serum fT4 declined during pregnancy and returned to baseline values post partum. TD in gestation was present in 23 women (7.9%): serum TSH was transiently decreased in 13 (6 had overt gestational thyrotoxicosis (2.1%)) and increased in 10 (2 had TPOAb). Both point prevalence and concentration of TPOAb decreased during gestation and returned to baseline levels within 12 weeks post partum. TD in post partum was present in 36 women (12.4%): 21 had subclinical and 15 overt TD. Out of the 15 women with overt TD (incidence of PPTD: 5.2%) 10 were positive for TPOAb (TPOAb+): 9 had thyrotoxicosis (4 TPOAb+), 5 hypothyroidism (5 TPOAb+) and 1 thyrotoxicosis followed by hypothyroidism (TPOAb+). Independent risk factors for PPTD were TPOAb (relative risk (RR) = 2 7.2), bottle feeding (RR = 11.1) and smoking habits (ever smoked: RR = 3.1; women with PPTD had smoked more cigarettes for a longer period of time). The sensitivity of TPOAb testing was highest at 12 weeks gestation (0.67). The ppv of TPOAb was 0.31-0.75 (depending on time of testing and concentration), increasing slightly to 0.38-0.80 when combined with bottle feeding or smoking habits. There appeared to be an autoimmune form of PPTD in 2/3 of cases and a non-autoimmune form; women with the autoimmune form were at risk for developing permanent hypothyroidism. CONCLUSIONS: A maximum of 2/3 of PPTD cases can be predicted from the presence of TPOAb because 1/3 remained negative for TPOAb. The most appropriate time for TPOAb testing is in the first trimester of pregnancy. The combination of TPOAb testing with anamnestic determinants of PPTD does not increase ppv substantially.

Cell-mediated immunity and postpartum thyroid dysfunction: a possibility for the prediction of disease?

Postpartum (pp) thyroid dysfunction (PPTD) is thought to be caused by an autoimmune (AI) destruction of thyroid follicles during the pp period. The chronic thyroid AI process [already present in pregnancy, as shown by the positivity for thyroid peroxidase antibodies (TPO-Ab)] becomes overt disease in the pp period, and one assumes that this exacerbation represents a rebound phenomenon after a general immunosuppression during pregnancy. The presence of TPO-Ab in pregnancy has been suggested as a predictor for later PPTD development. Apart from B cells, e.g. production of autoantibodies, various functions of the cell-mediated immune (CMI) system, including those of peripheral T cells, monocytes, and dendritic cells (DC), are also disturbed in AI states. The objectives of the present study were: determining alterations in various CMI parameters in pregnancies followed by PPTD vs. those not followed by PPTD; and determining the usefulness of these parameters in the prediction of PPTD. In a prospective study (region: Kempenland, southeast Netherlands), a random sample of 291 women were tested at 12 and 32 weeks gestation and 4 weeks pp for TPO-Ab. Women were followed until 9 months pp, for developing PPTD. PPTD was defined as both: an abnormal TSH, and fT4 pp women developing PPTD and/or being positive for TPO-Ab (n = 26); and thyroidological uneventful control women of the same cohort, matched for age and parity (n = 21), were tested for thyroid-stimulating antibodies, percentages of peripheral blood lymphocyte subsets using fluorescence-activated cell sorter analysis (CD3, CD4, CD8, CD16, CD56, major histocompatibility complex-class II), for monocyte polarization, and for cluster capability of monocyte-derived DC. Results were: 1) 31 women (10.7%) were positive for TPO-Ab (TPO-Ab+) in gestation (12 and/or 32 weeks); 2) 15 women (5.2%) developed PPTD, of whom 10 were TPO-Ab+ in gestation; 3) pregnancy-related CMI alterations consisted of low percentages of CD16+CD56+ natural killer (NK), cells and a low DC cluster capability at 12 weeks gestation (these functions were normalized at 32 weeks gestation); 4) the TPO-Ab+ PPTD+ women (4 hyper, 5 hypo, and 1 hyper/hypo) were characterized by a persistently low percentage of NK cells, a lowered monocyte polarization, and a raised percentage of major histocompatibility complex-class II+CD3+ T cells; 5) the TPO-Ab- PPTD+ women (all 5 hyper) had neither thyroid-stimulating antibodies nor CMI alterations, apart from those normally seen in pregnancy; 6) 21 women were positive for TPO-Ab in pregnancy but did not develop PPTD (they had the same lowered NK cell percentages and monocyte polarization as the TPO-Ab+ PPTD+ cases, but they had normal percentages of activated peripheral T cells and a lower titer of TPO-Ab); 7) determination of the number of NK cells and monocyte polarization hardly contributed to the prediction of PPTD (as compared with TPO-Ab status), because of strong interindividual variation and close association with the presence of TPO-Ab; and 8) combining TPO-Ab assays with testing for activated T cells was the most optimal parameter for the prediction of TPO-Ab+ cases of PPTD in our small test set. We conclude that TPO-Ab+ pregnant women who develop PPTD show several CMI abnormalities other than those seen in normal pregnant women, such as persistently lower percentage of NK cells, a lowered monocyte polarization, and a raised percentage of activated T cells. The latter seems rather specific for the actual PPTD development and is not found in TPO-Ab+ (but PPTD) uncomplicated pregnancies. TPO-Ab- (but PPTD+) women had no signs of CMI abnormalities (apart from those specific for the pregnancy state). Although studied cases are low in number, our data are suggestive for the existence of two forms of PPTD: a TPO-Ab+ (AI) form (two-thirds of patients, classical PPTD pattern); and a TPO-Ab- (non-AI) form (one-third of patients, only hyper). Such assumption implies that, at best, two

[Thyroid cancer in Southeastern Netherlands, 1970-1989: trends in incidence, treatment and survival]. / Schildklierkanker in Zuidoost-Nederland, 1970-1989: trends in incidentie, behandeling en overleving.

OBJECTIVE: To describe the (changes in) incidence, treatment and prognosis of thyroid cancer (TC) in the period 1970-89 in the South-east of the Netherlands. SETTING: Eindhoven Cancer Registry, Comprehensive Cancer Centre South (I.K.Z.), Eindhoven. DESIGN: Retrospective. PATIENTS AND METHODS: Data were collected from al TC patients (ICD-O code 193 and non-Hodgkin lymphoma originating from the thyroid) diagnosed in the period 01.01.1970-31.12.1989. Histological, treatment and survival (on 01.07.1991) data were collected. The standardised incidence, prevalence, (relative) survival and mortality were calculated for men and women. RESULTS: The mean age of the 297 TC patients was 52 years; the male-female ratio was 1:2.3. 46% Of the TC patients had papillary TC, 35% follicular TC. The incidence of TC increased from 1.4 to 3.0/100,000/yr for females, the incidence remained unchanged for males (1,1/100,000/yr). The point prevalence (01.07.1991) was 25.6/100,000 for females and 8.2/100.000 for males. TC patients were treated in all hospitals in the region and were increasingly first seen bij an internist. There was a trend to more extensive surgical treatment and iodine-131 treatment in differentiated TC. For all TC patients the crude 10-year survival rate was 61%, the relative 10-year survival rate was 74%. Survival was related with sex, age and histological type. Mortality from TC remained very low. CONCLUSIONS: The incidence and prognosis of TC were similar to the surrounding countries. In general the recommendations from the consensus meetings for treatment of TC in 1985 and 1987 appear to be followed.

[Cost-effectiveness of lowering of blood cholesterol using simvastatin and cholestyramine]. / De kosten-effectiviteit van verlaging van het cholesterolgehalte met simvastatine en colestyramine.

With use of a model of the costs and effects of cholesterol lowering therapy in the primary prevention of coronary heart disease, the cost-effectiveness of simvastatin and cholestyramine in the Netherlands have been estimated. Costs per year of life saved by cholestyramine therapy are several times greater than those of simvastatin therapy and compared unfavorably with those of generally accepted health care programs in the Netherlands. Cholesterol-lowering with simvastatin in men can be cost-effective when therapy is initiated at an early age. At cholesterol levels between 6.5 and 8 mmol/l, however, therapy should be restricted to men with at least one, preferably two additional risk factors such as hypertension or diabetes mellitus. Among women, cholesterol lowering can only be cost-effective when therapy is limited to women with diabetes mellitus or severely elevated serum cholesterol levels.