Weight status and nonadherence to asthma maintenance therapy among children enrolled in a public drug insurance plan.

Objective: The pediatric obese-asthma phenotype is associated with poor control, perhaps because of medication nonadherence. This study aimed to assess whether weight status is associated with nonadherence in children prescribed new asthma maintenance therapies.Methods: A historical cohort was constructed from a clinical database linking individual patient and prescription data to Quebec's prescription claims registry. Children aged 2-18 years with specialist-diagnosed asthma who were newly prescribed one of the following maintenance controllers: leukotriene receptor antagonists (LTRA); low-dose inhaled corticosteroids (ICS); medium/high-dose ICS; or combination therapy (ICS with long-acting beta-2 agonists and/or LTRA), at the Asthma Center of the Montreal Children's Hospital from 2000-2007 were included. Primary nonadherence was defined as not claiming any prescriptions, whereas secondary nonadherence was measured with the proportion of prescribed days covered (PPDC ≤ 50%) among primary adherers over a 6-month follow-up period. A modified Poisson regression model served to estimate the effect of excess weight (BMI > 85th percentile) on primary and secondary nonadherence.Results: Approximately one third of patients were primary nonadherers and 60% took less than 50% of prescribed therapy. Excess weight was associated with a trend toward increased risk of primary nonadherence in children newly prescribed low-dose ICS (RR 1.53, 95%CI 0.94-2.49), and of secondary nonadherence in children initiating medium/high-dose ICS (RR 1.24; 95%CI 0.98-1.59).Conclusions: Excess weight status is a possible determinant of primary nonadherence in children initiating low-dose ICS and secondary nonadherence to higher-dose ICS regimens. This hypothesis-generating study suggests that nonadherence may be a potential contributor to higher morbidity in children with obese-asthma.

The obese-asthma phenotype in children: An exacerbating situation?

BACKGROUND: Current evidence regarding the relationship between childhood obesity, decreased response to inhaled corticosteroids (ICSs), and poor asthma control is conflicting. OBJECTIVES: We assessed whether obesity (1) is associated with time to first exacerbation among children with asthma initiating step 3 maintenance therapies and (2) modifies the effectiveness of step 3 therapies. METHODS: A retrospective cohort study was conducted from clinical data linked to health and drug administrative databases. The cohort consisted of children aged 2 to 18 years with specialist-confirmed asthma who initiated medium/high-dose ICS monotherapy or low/medium-dose ICS with leukotriene receptor antagonist/long-acting ß-agonist (combination therapy) at the Montreal Children's Hospital Asthma Center from 2000 to 2007. Children were classified as exposed to step 3 therapies when they were dispensed a corresponding drug claim during follow-up, whereas those without claims were classified as nonadherers. Marginal structural Cox models were used to estimate the effect of obesity (body mass index > 97th percentile) and treatment on time to exacerbation, which was defined as any emergency department visit, hospitalization, or use of oral corticosteroids for asthma. RESULTS: Of the 4621 cohort patients, 231 initiated ICS monotherapy, and 97 initiated combination therapy. The hazard ratio (HR) for obesity was 1.67 (95% CI, 1.41-1.98). Compared with nonobese nonadherers, the HR for obese nonadherers was 1.54 (95% CI, 0.97-2.45); the HR for ICS monotherapy in obese and nonobese children was 0.85 (95% CI, 0.47-1.52) and 0.58 (95% CI, 0.37-0.91), respectively; and the HR for combination therapy in obese and nonobese children was 0.50 (95% CI, 0.13-1.89) and 0.46 (95% CI, 0.23-0.92), respectively. CONCLUSION: Obesity might be a determinant of shorter exacerbation-free time in children with asthma; however, we could not rule out a differential response to step 3 therapies by obesity status, potentially because of a lack of precision.

The Experience of Patients Engaged in Co-designing Care Processes.

This article presents the experiences of patients engaged in co-designing care under a program entitled, "Transforming Care at the Bedside," based at an academic health sciences center. This descriptive, qualitative study collected data through individual interviews. Participants included patients from 5 units in an academic health sciences center in Quebec, Canada. A total of 6 individual interviews were conducted in November 2014, 15 months after the Transforming Care at the Bedside work began in September 2013. Content analysis was used to analyze the qualitative data. Being listened to and informed gave patients an opportunity to better understand patient needs and the complexity of care in the unit and in the organization. The experience enabled patients to better translate the patient experience for the team's benefit and influence the team's perspective and decisions. Through this experience, several patients felt motivated and empowered and that they afforded consideration through this experience. This study highlights the importance of creating opportunities for patients and health care providers to share their unique experiences and expertise to better understand each other's reality. In this context, they developed a more comprehensive understanding of the issues and worked together to implement realistic changes on behalf of the patients.

Effective Strategies to Spread Redesigning Care Processes Among Healthcare Teams.

PURPOSE: The purpose of this study was to describe how spread strategies facilitate the successful implementation of the Transforming Care at the Bedside (TCAB) program and their impact on healthcare workers and patients in a major Canadian healthcare organization. DESIGN: This study used a qualitative and descriptive design with focus groups and individual interviews held in May 2014. Participants included managers and healthcare providers from eight TCAB units in a university health center in Quebec, Canada. The sample was composed of 43 individuals. METHODS: The data were analyzed using NVivo according to the method proposed by Miles and Huberman. FINDINGS: The first two themes that emerged from the analysis are related to context (organizational transition requiring many changes) and spread strategies for the TCAB program (senior management support, release time and facilitation, rotation of team members, learning from previous TCAB teams, and engaging patients). The last theme that emerged from the analysis is the impact on healthcare professionals (providing front-line staff and managers with the training they need to make changes, team leadership, and increasing receptivity to hearing patients' and families' needs and requests). CONCLUSIONS: This study describes the perspectives of managers and team members to provide a better understanding of how spread strategies can facilitate the successful implementation of the TCAB program in a Canadian healthcare organization. CLINICAL RELEVANCE: Spread strategies facilitate the implementation of changes to improve the quality and safety of care provided to patients.

Transforming Care at the Bedside: managers' and health care providers' perceptions of their change capacities.

BACKGROUND: The Transforming Care at the Bedside (TCAB) project engages frontline health care providers as the leaders of change and improvement efforts in their work environment. This study explored how health care providers and managers from three TCAB units in a university-affiliated health care center perceived the development of their change capacities following their involvement in this program. METHOD: This descriptive, qualitative study involved focus groups and individual interviews. RESULTS: Participants learned to work as a team and to expand their outlook. They had access to processes and tools to learn new skills. New relationships also developed among the various players, and they shared new roles, which enabled them to translate the desired changes into action and make the results visible. CONCLUSION: The study showed the TCAB program helps develop health care providers' and managers' change capacities.

The effect of transforming care at the bedside initiative on healthcare teams' work environments.

BACKGROUND: Different initiatives have been implemented in healthcare organizations to improve efficiency, such as transforming care at the bedside (TCAB). However, there are important gaps in understanding the effect of TCAB on healthcare teams' work environments. AIM: The specific aim of the study is to describe findings regarding the TCAB initiative effects on healthcare teams' work environments. METHODS: A pretest and posttest study design was used for this study. The TCAB initiative was implemented in fall 2010 in a university health center in Montreal, Canada. The sample consisted of healthcare workers from four different care units. RESULTS: Statistically significant improvement was observed with the communicating specific information subscale from the measure of processes of care variable, and a significant difference was found between the support from colleagues variable, which was higher at baseline than postprogram. The differences for psychological demand, decisional latitude, and effort-reward were not significant. CONCLUSIONS: TCAB is an intervention that allows healthcare teams to implement change to improve patients' and families' outcomes. Ongoing energy should focus on how to improve communication among all members of the team and ensure their support.

The effect of obesity on antibiotic treatment failure: a historical cohort study.

PURPOSE: Obesity, a major health issue, is also an important risk factor for infections. Evidence demonstrates that excess weight affects the disposition of antibiotics but little work has been done to explore if this results in antibiotic treatment failure (ATF). ATF has serious adverse health outcomes and may increase treatment resistance. Given that obese patients often have other health issues, it is important to determine if excess weight independently increases the likelihood of ATF. METHODS: Consenting patients (N = 18 014), randomly sampled from Santé Québec Health surveys (1992, 1998), were linked with administrative health databases. Patients were within the normal, overweight, and obese weight categories aged 20-79 years old, receiving at least one course of antibiotic therapy from the survey date until December 2005. ATF was defined as any additional antibiotic prescriptions or hospitalizations for infections within the 30 days after initial therapy. Logistic regression was used to assess the impact of excess weight on ATF after adjusting for patient characteristics, comorbidities, history of antibiotic use, antibiotic resistance, and flu season. RESULTS: Of the final sample size (N = 6 179), 39.0% were overweight and 21.4% were obese. The most frequently prescribed antibiotics were amoxicillin (16.0%), ciprofloxacin (9.2%), phenoxymethylpenicillin (8.8%), trimethroprim/sulfamethoxazole (8.6%), and clarithromycin (8.5%). ATF occurred in 828 (13.4%) of the 6 179 study patients. Obesity was a significant predictor of ATF (adjusted OR 1.26; 95% CI 1.03-1.52). CONCLUSION: Obesity is a significant risk factor for ATF, and this association may be due to the current "one size fits all" dosing strategy, which warrants further investigation.

Influence of weight status in the response to Step-2 maintenance therapies in children with asthma.

Introduction: Overweight children with asthma may display impaired response to inhaled corticosteroids (ICS), possibly due to non-eosinophilic inflammation or weight-related lung compression; these mechanisms may differentially affect response to ICS and leukotriene receptor antagonists (LTRAs). We assessed whether weight status modified the response to low-dose ICS and LTRA Step-2 monotherapy. Methods: A historical cohort study from clinical data linked to administrative databases was conducted among children aged 2-18 years with specialist-diagnosed asthma who were initiating or continuing a Step-2 monotherapy from 2000 to 2007 at the Montreal Children's Hospital Asthma Centre. The outcome was time-to-management failure defined as any step-up in therapy, acute care visit, hospitalisation or oral corticosteroids for asthma, whichever occurred first. The independent and joint effects of weight status (body mass index [BMI] percentile) and time-varying treatment on time-to-management failure were estimated with marginal structural Cox models. The likelihood ratio test (LRT) and relative excess risk due to interaction (RERI) were computed to assess treatment effect modification by weight status on the multiplicative and additive scales. Results: Of the 433 and 85 visits with a low-dose ICS and LTRA prescription, respectively, 388 management failures occurred over 14 529 visit-weeks of follow-up. Children using LTRA compared with low-dose ICS tended to have an overall higher risk of early management failure (HR 1.52; 95% CI 0.72 to 3.22). Irrespective of treatment, the hazard of management failure increased by 5% (HR 1.05; 95% CI 1.01 to 1.10) for every 10-unit increase in BMI percentile. An additional hazard reduction of 17% (HR 0.83; 95% CI 0.70 to 0.99) was observed for every 10-unit increase in BMI percentile among LTRA users, but not for ICS (HR 0.95; 95% CI 0.86 to 1.04). The LRT indicated a departure from exact multiplicativity (p<0.0001), and the RERIs for ICS and LTRA were -0.05 (95% CI -0.14 to 0.05) and -0.52 (95% CI -1.76 to 0.71). Conclusions: Weight status was associated with earlier time-to-management failure in children prescribed Step-2 therapy. This hypothesis-generating study suggests that LTRA response increases in children with higher BMI percentiles, although further research is warranted to confirm findings.

Impact of patient communication problems on the risk of preventable adverse events in acute care settings.

BACKGROUND: Up to 50% of adverse events that occur in hospitals are preventable. Language barriers and disabilities that affect communication have been shown to decrease quality of care. We sought to assess whether communication problems are associated with an increased risk of preventable adverse events. METHODS: We randomly selected 20 general hospitals in the province of Quebec with at least 1500 annual admissions. Of the 145,672 admissions to the selected hospitals in 2000/01, we randomly selected and reviewed 2355 charts of patients aged 18 years or older. Reviewers abstracted patient characteristics, including communication problems, and details of hospital admission, and assessed the cause and preventability of identified adverse events. The primary outcome was adverse events. RESULTS: Of 217 adverse events, 63 (29%) were judged to be preventable, for an overall population rate of 2.7% (95% confidence interval [CI] 2.1%-3.4%). We found that patients with preventable adverse events were significantly more likely than those without such events to have a communication problem (odds ratio [OR] 3.00; 95% CI 1.43-6.27) or a psychiatric disorder (OR 2.35; 95% CI 1.09-5.05). Patients who were admitted urgently were significantly more likely than patients whose admissions were elective to experience an event (OR 1.64, 95% CI 1.07-2.52). Preventable adverse events were mainly due to drug errors (40%) or poor clinical management (32%). We found that patients with communication problems were more likely than patients without these problems to experience multiple preventable adverse events (46% v. 20%; p = 0.05). INTERPRETATION: Patients with communication problems appeared to be at highest risk for preventable adverse events. Interventions to reduce the risk for these patients need to be developed and evaluated.

The Impact of Antidepressant Therapy on Glycemic Control in Canadian Primary Care Patients With Diabetes Mellitus.

Context: Depression is common in people with diabetes and is associated with poor glycemic control. Evidence suggests that certain antidepressants (AD) increase the risk of poor control. Few population-based studies have examined the impact of individual ADs on glycemic control. This study's objective is to measure the impact of Citalopram, Amitriptyline, Venlafaxine, Trazodone and Escitalopram on glycated hemoglobin (HbA1c) in Canadian primary care patients with diabetes. Methods: A retrospective study of electronic medical records (EMR) from 115 primary care practices across Canada was undertaken. Data were obtained from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN). The sample population comprised 1,084 diabetic patients with 1,127 prescriptions of one of the five selected ADs and with baseline and post-exposure HbA1c measurements. Generalized linear mixed models were computed to estimate the effect of the ADs on HbA1c. Results: Mean HbA1c ratios for Amitriptyline, Venlafaxine, Trazodone and Escitalopram were all numerically lower than Citalopram. The confidence intervals included the minimum detectable effect, however the differences were not statistically significant. The lowest clinically relevant HbA1c ratios, relative to Citalopram, were found in patients prescribed Trazodone and Escitalopram. Accounting for the prescription of Trazodone for indications other than depression, this research suggests that Escitalopram may be safer than Citalopram for people with diabetes and depression, in terms of its effect on blood glucose. Conclusion: This study can inform future research examining the relationship between ADs and blood glucose and provides insight into the limitations pertaining to the use of health data in health research. Future research should seek to control for, across multiple time points: depression symptoms, depression severity, depression duration, weight, diabetes medication, tobacco and alcohol consumption and other medications with a known impact on blood glucose.

The Importance of Relevance: Willingness to Share eHealth Data for Family Medicine Research.

Objective: To determine the proportion of family medicine patients unwilling to allow their eHealth data to be used for research purposes, and evaluate how patient characteristics and the relevance of research impact that decision. Design: Cross-sectional questionnaire. Setting: Acute care respiratory clinic or an outpatient family medicine clinic in Montreal, Quebec. Participants: Four hundred seventy-four waiting room patients recruited via convenience sampling. Main Outcome Measures: A self-administered questionnaire collected data on age, gender, employment status, education, mother tongue and perceived health status. The main outcome of was self-reported relevance of three research scenarios and willingness or refusal to share their anonymized data. Responses were compared for family practice vs. specialty care patients. Results: The questionnaire was completed by 229 family medicine respondents and 245 outpatient respondents. Almost a quarter of all respondents felt the research was not relevant. Family medicine patients (15.7%) were unwilling to allow their data to be used for at least one scenario vs. 9.4% in the outpatient clinic. Lack of relevance (OR 11.55; 95% CI 5.12-26.09) and being in family practice (OR 2.13; 95% CI 1.06-4.27) increased the likelihood of refusal to share data for research. Conclusion: Family medicine patients were somewhat less willing to share eHealth data, but the overall refusal rate indicates a need to better engage patients in understanding the significance of full access to eHealth data for the purposes of research. Personal relevance of the research had a strong impact on the responses arguing for better efforts to make research more pertinent to patients.

Creating non-parametric bootstrap samples using Poisson frequencies.

This article describes how, in the high-level software packages used by non-statisticians, approximate non-parametric bootstrap samples can be created and analyzed without physically creating new data sets, or resorting to complex programming. The comparable performance of this shortcut method, which uses Poisson rather than multinomial frequencies for the numbers of copies of each observation, is demonstrated theoretically by evaluating the bootstrap variance in an example where the classic estimator of the sampling variance of the statistic of interest has a known closed form. For sample sizes of 50 or more, bootstrap standard errors obtained by this shortcut method exceeded those obtained by the standard version by less than 1%. The proposed method is also evaluated in two worked examples, involving statistics whose sampling distribution is more complex. The second of these is also used to illustrate when one can and cannot use non-parametric bootstrap samples.

Studies of prognostic factors in end-stage renal disease: an epidemiological and statistical critique.

We reviewed prognostic studies for patients treated with renal replacement therapy by using an electronic database and bibliographic review for 1990 to 1998. Using the inclusion criteria of English language, adult patients, primary article, minimum 50 patients, primary focus on prognostic factors, and mortality outcome, 104 articles were identified. The 104 articles were reviewed for eight epidemiological and seven statistical criteria that addressed the scientific validity and interpretability of results. The following percentages of the 104 articles satisfied each of the eight epidemiological criteria: (1) a priori hypothesis, 6%; (2) zero time specified, 49%; (3) prognostic factors collected before zero time, 69%; (4) inception cohort, 59%; (5) control for treatment, 74%; (6) operational criteria, 82%; (7) missing variables reported, 12%; and (8) loss to follow-up reported, 42%. Summary analysis showed that 76% of studies satisfied four or fewer of the eight identified criteria. In the 77 articles (74%) that used the Cox proportional hazards model, the following percentages of articles met each of the seven statistical criteria: (1) proportional hazards verified, 26%; (2) censoring explained, 57%; (3) multivariate analysis performed, 91%; (4) significance levels given, 99%; (5) age adjusted, 95%; (6) diabetes adjusted, 66%; and (7) cardiac adjusted, 44%. Summary analysis found that 47% of the 77 studies satisfied four or fewer of the seven identified criteria. Superficially, results appear to show that when the Cox proportional hazards model was used, statistical analysis was better than the epidemiological design. However, studies we examined had serious defects in both epidemiological design and statistical analysis. The consequent validity of results for the quantification of prognostic factors is questionable.

Quantifying postural tremor in workers exposed to low levels of manganese.

The aim of this study was: (1) To determine the minimum number of characteristics necessary to discriminate between postural tremor recorded in control subjects (CO), in subjects exposed to manganese (MN), and in patients with Parkinson's disease (PD), and (2) to examine the continuum of changes between the three groups examined. Workers previously exposed to Mn (n = 10), patients with PD (n = 10), and control subjects (CO) (n = 11) underwent a clinical examination. Blood Mn was measured at the end of exposure time for the MN group and 12 months later at the beginning of the experiment for all groups. Postural tremor with visual feedback was recorded in the index finger with a laser system. Statistical criteria were used to reduce computed tremor characteristics to a minimal set of reliable discriminating variables. Two variables were retained namely corrected wobble (CW), describing the morphology of the tremor oscillations, and variability ratio (VR), describing proportional power of tremor. Both variables had an overall correct classification rate of 77.4%. Blood Mn levels at the time of the experiment were similar for all groups and had insignificant correlation with tremor variables. However, blood Mn levels in workers which were also measured at the end of exposure time (i.e., 12 months before) showed significant correlation (Spearman's rank coefficient) with both harmonic index (rho = 0.70, P = 0.03) and first maximum of the autocorrelation function (rho = 0.89, P = 0.001). We conclude that (1) the tremor of workers exposed to Mn could be adequately described with only two variables; (2) a continuum of changes between tremor recorded in control subjects, in subjects exposed to Mn and in patients with PD was observed, with the MN group always found in between the control (CO) and the PD groups; (3) while blood Mn levels in workers were back at control levels at the time of the experiment, the effect of Mn on postural tremor was still detected. Thus our method has the potential to detect the effect of Mn on tremor with only two variables even after Mn level in the blood is back to normal values.

Blood manganese and alcohol consumption interact on mood states among manganese alloy production workers.

Long-term exposure to manganese (Mn) can induce neurotoxic effects including neuromotor, neurocognitive and neuropsychiatric effects, but there is a great interpersonal variability in the occurrence of these effects. It has recently been suggested that blood Mn (MnB) may interact with alcohol use disorders, accentuating neuropsychiatric symptoms. The objective of the present study was to explore a possible interaction between alcohol consumption and MnB on mood states, using an existing data set on Mn exposed workers. Respirable Mn exposure in the plant averaged 0.23mg/m(3) and was correlated with MnB. All participants for whom all data on MnB concentration and mood (assessed with the Profile of Mood States (POMS)) were available and who reported currently drinking alcohol were included in the analyses (n=74). Workers were grouped according to their MnB concentration (<10 and > or =10 microg/l) and alcohol consumption (<400 and > or =400g per week). Two-way ANOVAs were performed on each POMS scale and Mann-Whitney tests were used to assess group differences. Workers in the higher alcohol consumption group had higher scores on three POMS scales: tension, anger and fatigue. There was no difference for POMS scale scores between MnB subgroups. Dividing the group with respect to alcohol consumption and MnB showed that the group with high alcohol consumption and high MnB displayed the highest scores. In the lower MnB category, those in the higher alcohol consumption group did not have higher scores than the others. The interaction term for alcohol consumption and MnB concentration was statistically significant (P<0.05) for the depression, anger, fatigue and confusion POMS scales. There was a tendency for tension (P<0.06), and it was not significant for vigor. This study shows the first evidence of an interaction between MnB and alcohol consumption on mood states among Mn exposed workers and supports the results from a previous population-based study.

CE: Original research: the perceptions of health care team members about engaging patients in care redesign.

OBJECTIVE: This study sought to explore the perceptions of health care workers about engaging patients as partners on care redesign teams under a program called Transforming Care at the Bedside (TCAB), and to examine the facilitating factors, barriers, and effects of such engagement. DESIGN: This descriptive, qualitative study collected data through focus groups and individual interviews. Participants included health care providers and managers from five units at three hospitals in a university-affiliated health care center in Canada. METHODS: A total of nine focus groups and 13 individual interviews were conducted in April 2012, 18 months after the TCAB program began in September 2010. Content analysis was used to analyze the qualitative data. FINDINGS: Health care providers and managers benefited from engaging patients in the decision-making process because the patients brought a new point of view. Involving the patients exposed team members to valuable information that they hadn't previously thought about during decision making. CONCLUSION: Health care teams stand to benefit from engaging patients in the change process. Patients contribute a different point of view, and this helps to ensure that the changes proposed and implemented address their needs.

Nipple aspirate fluid and ductoscopy to detect breast cancer.

We prospectively performed cytologic assessment and image analysis (IA) on matched nipple aspirate fluid (NAF) and mammary ductoscopy (MD) specimens to determine (1) the accuracy of these methods in cancer detection and (2) whether the two collection methods provide complementary information.NAF and MD specimens were collected from 84 breasts from 75 women (nine bilateral samples) who underwent breast surgery. Cytologic evaluation was performed on all samples. IA was performed on slides with sufficient epithelial cells.Cytologic evaluation proved more accurate in patients without pathologic spontaneous nipple discharge (PND) than those with PND, mainly because of the potential false positive diagnosis in the latter. While the sensitivity of NAF and MD cytology was low (10% and 14%, respectively), both were 100% specific in cancer detection in the non-PND cohort. Combining NAF and MD cytology information improved sensitivity (24%) without sacrificing specificity. Similar to cytology, IA was more accurate in patients without PND having high specificity (100% for aneuploid IA), but relatively low sensitivity (36%). Combining NAF and MD cytology with aneuploid IA improved the sensitivity (45%) while maintaining high specificity (100%). The best predictive model was positive NAF cytology and/or MD cytology combined with IA aneuploidy, which resulted in 55% sensitivity and 100% specificity in breast cancer detection.Cytologic evaluation and IA of NAF and MD specimens are complementary. The presence of atypical cells arising from an intraductal papilloma in ductoscopic specimens is a potential source of false positive diagnosis in patients with nipple discharge.

Outlier detection for a hierarchical Bayes model in a study of hospital variation in surgical procedures.

One of the most important aspects of profiling healthcare providers or services is constructing a model that is flexible enough to allow for random variation. At the same time, we wish to identify those institutions that clearly deviate from the usual standard of care. Here, we propose a hierarchical Bayes model to study the choice of surgical procedure for rectal cancer using data previously analysed by Simons et al.(1) Using hospitals as random effects, we construct a computationally simple graphical method for determining hospitals that are outliers; that is, they differ significantly from other hospitals of the same type in terms of surgical choice.

Sampling for conditional inference on case-control data.

The problem of exact conditional inference for discrete multivariate case-control data has two forms. The first is grouped case-control data, where Monte Carlo computations can be done using the importance sampling method of Booth and Butler (1999, Biometrika86, 321-332), or a proposed alternative sequential importance sampling method. The second form is matched case-control data. For this analysis we propose a new exact sampling method based on the conditional-Poisson distribution for conditional testing with one binary and one integral ordered covariate. This method makes computations on data sets with large numbers of matched sets fast and accurate. We provide detailed derivation of the constraints and conditional distributions for conditional inference on grouped and matched data. The methods are illustrated on several new and old data sets.

Fiberoptic ductoscopy findings in women with and without spontaneous nipple discharge.

BACKGROUND: Fiberoptic ductoscopy (FD), which allows direct visualization of the breast ductal lumen, is performed in women with and without spontaneous nipple discharge (SND). Previous reports suggested that cytologic evaluation of SND may be falsely interpreted as containing malignancy. The purpose of the current study, which was performed prospectively, was to determine whether ductoscopic findings were different in women with versus without SND, and to assess the implications of the differences in SND versus non-SND samples regarding the role of FD in assessing whether a woman has breast carcinoma. METHODS: Data were collected on the distance traveled by the ductoscope, visual observations, pathology, cytology, epithelial and foam cell quantity, and image analysis for ploidy, hypertetraploidy, and S-phase fraction. RESULTS: Of 100 FD specimens, 60 were from breasts without SND and 40 were from breasts with SND. Intraductal visual observations (P < or = 0.0002), pathologic findings in the resected specimen (P < or = 0.001), and quantity of epithelial cells (P=0.03) were influenced by the presence or absence of SND. Although one specimen from a benign breast was interpreted as cytologically malignant, every breast with both malignant cytology and aneuploidy contained cancer cells. A model incorporating cytology and SND was 92% sensitive and 60% specific in predicting which women had breast carcinoma. CONCLUSIONS: There were pronounced differences in FD samples from women with and without SND. FD biologic parameters can be chosen to optimize breast carcinoma predictive sensitivity and specificity. SND cytology can present a diagnostic problem, suggesting the need for histologic confirmation before the initiation of therapy.