Impact of sex, age, race, ethnicity and aspirin use on bleeding symptoms in healthy adults

Journal of Thrombosis and Haemostasis, 9: 100–108 Impact of sex, age, race, ethnicity and aspirin use on bleedingsymptoms in healthy adults A . C . M A U E R , * N . A . K H A Z A N O V , * N . L E V E N K O V A , * S . T I A N , * E . M . B A R B O U R , * C . K H A L I D A ,  J . N . T O B I N *   and B . S . C O L L E R **The Rockefeller University, New York, NY; and  Clinical Directors Network, Inc. (CDN), New York, NY, USA To cite this article: Mauer AC, Khazanov NA, Levenkova N, Tian S, Barbour EM, Khalida C, Tobin JN, Coller BS. Impact of sex, age, race, ethnicity and aspirin use on bleeding symptoms in healthy adults. J Thromb Haemost 2011; 9: 100–8.
Keywords: bleeding disorders, bleeding history, gender differ- Summary.Background: Comparing a patientÕs bleeding symp- toms with those of healthy individuals is an importantcomponent of the diagnosis of bleeding disorders, but little is known about whether bleeding symptoms in healthy individualsvary by sex, race, ethnicity, age, or aspirin use. Objectives, Obtaining a detailed bleeding history is an important compo- Patients/Methods: We developed a comprehensive, ontology- nent of the medical evaluation to determine whether a person: backed, Web-based questionnaire to collect bleeding histories (i) has a bleeding diathesis; (ii) is at increased risk of excessive from 500 healthy adults. The mean age was 43 years (range 19– hemorrhage in response to invasive procedures and/or surgery; 86 years), 63% were female, 19% were Hispanic, 37% were and/or (iii) should undergo laboratory evaluation and/or African-American, 43% were Caucasian, 8% were Asian, and referral to a specialist [1–3]. Moreover, correlation of clinical 4% were multiracial. Results: 18 of the 36 symptoms captured hemorrhage with genetic, biochemical and/or functional data occurred with < 5% frequency, and 26% of participants related to platelets or coagulation factors can provide impor- reported no bleeding symptoms (range 0–19 symptoms).
tant information for a better understanding of the basic Differences in sex, race, ethnicity, aspirin use and age accounted mechanisms of hemostasis. However, despite a number of early for only 6–13% of the variability in symptoms. Although men attempts to standardize bleeding history questionnaires (re- reported fewer symptoms than women (median 1 vs. 2, viewed by Coller and Schneiderman [1]), no generally recog- P < 0.01), there was no difference when sex-specific questions nized standards for collecting a bleeding history emerged until were excluded (median 1 for both men and women, P = 0.50).
recently, when Rodeghiero et al. developed bleeding question- However, women reported more easy bruising (24% vs. 7%, naires and scoring systems for the evaluation of patients with P < 0.01) and venipuncture-related bruising (10% vs. 3%, von Willebrand disease (VWD). By assigning point values to P = 0.02). The number of symptoms did not vary by race or the severity of various bleeding symptoms, they demonstrated age, but epistaxis was reported more frequently by Caucasians that a bleeding score is valuable for confirming the diagnosis of than by African-Americans (29% vs. 18%, P = 0.02), and VWD [4–6] and predicting the risk of future hemorrhagic epistaxis frequency decreased with age (odds ratio 0.97 per year, events in type 1 [5] and type 2B [7] VWD. Several groups have P < 0.01). Paradoxically, infrequent aspirin users reported used the Vicenza bleeding score as is or with modifications to more bruising and heavy menses than frequent users (21% vs.
8%, P = 0.01, and 56% vs. 38%, P = 0.03, respectively).
The value of such approaches depends on comparing the Conclusions: Our findings provide a contemporaneous and bleeding histories of affected and healthy individuals. The comprehensive description of bleeding symptoms in a diverse literature, however, reports marked variability of hemorrhagic group of healthy individuals. Our Web-based system is freely symptoms among apparently healthy individuals (Table 1).
For instance, reported symptom frequencies range from 2% to85% for epistaxis, from 20% to 44% for menorrhagia, from0% to 11% for postoperative bleeding, from 0% to 35% forbleeding after tooth extraction, and from 11% to 61% for Correspondence: Andreas C. Mauer, 1230 York Ave, Box 309, New gingival bleeding [11–17]. At present, it is unclear whether these variations reflect differences in the questionnaires used, the Tel.: +1 212 327 7633; fax: +1 212 327 7493.
methods of administration, the sex, race, ethnicity, age, or frequency of use of medications with antiplatelet effects of the Received 20 August 2010, accepted 5 October 2010 populations sampled, or other factors.
Ó 2010 International Society on Thrombosis and Haemostasis Bleeding symptoms in healthy individuals 101 Although only women face the hemostatic challenges of menstruation, pregnancy, and childbirth, there is only limited information on whether other bleeding symptoms differ between men and women. We found only a single study, which was limited to the evaluation of differences in epistaxis frequency between Caucasians and Asians [18], that assessed whether individuals of different racial backgrounds experience bleeding symptoms with different frequencies. As Hispanic ethnicity is viewed as distinct from race [19], we also searched for, but did not find, reported differences in bleeding symptoms between individuals of Hispanic and non-Hispanic ethnicity.
Although it is recognized that some bleeding symptoms such as epistaxis occur more frequently in childhood than adulthood [1], we could not identify information on whether bleedingsymptoms differ among adults of different ages. Moreover, as older individuals have had more time to sustain hemorrhagic symptoms, it is possible that age is an important factor in interpreting the bleeding history. Aspirin has significant antiplatelet properties, but the association between aspirin use and bleeding symptoms in a contemporaneous healthy population has not been systematically assessed with a To address these questions, we designed and deployed a comprehensive, Web-based bleeding history questionnaire in a study that aimed to: (i) establish the frequencies of bleeding symptoms in a diverse population of healthy adults; and (ii) assess whether any symptoms varied by sex, race, ethnicity, age, The design of the bleeding history phenotyping system we used has been previously described [20]. The centerpiece is a comprehensive bleeding history questionnaire that incorpo- rated elements from a review of the literature, one of the authorsÕ experience [1], and input from experts in hemostasis, questionnaire development, epidemiology, and biomedical informatics. It contains 278 questions covering 25 categories of bleeding and related covariates, with the latter including dermatologic lesions, connective tissue disorders, medications, and family history. To reduce ambiguity in terminology, 168 terms in the questionnaire and ontology were cross-referenced to the National Library of MedicineÕs Unified Medical Language System by assigning them the corresponding code The questionnaire was used to derive a bleeding history ontology, which is an explicit representation of the relation- ships among bleeding signs, symptoms, disorders, and treat- ments. The ontology formalizes the concepts contained in the questionnaire in an electronic format that facilitates data analysis, organization, and representation. The ontology is publicly available in the Bioportal ontology registry (http:// bioportal.bioontology.org/ontologies/40546) to facilitate its critique by experts in the field and its future updating, as, for Ó 2010 International Society on Thrombosis and Haemostasis example, when new therapies for bleeding disorders are 235 individuals were approached and 135 (57%) agreed to participate. In both settings, after written informed consent was The questionnaire is administered by a medically trained obtained, a physician or nurse trained in the use of the individual using a Web-based program. Studies are assigned a questionnaire conducted the interview and entered the partic- site identification code as well as a code to identify the person ipantÕs responses directly into the database with a personal administering the questionnaire. To ensure confidentiality, computer. Personnel conducting the interviews completed a each respondent is identified by a randomly generated unique credentialing process that included: (i) observing one of us personal identification number. Skip patterns were introduced (ACM) conduct two interviews; (ii) obtaining two histories into the program to speed questionnaire completion. For under ACMÕs direct supervision; and (iii) completing a example, respondents who state that they have never had checklist of skills. ACM conducted 129 interviews, trained epistaxis are not asked questions about epistaxis frequency or research nurses conducted 236 interviews, and CDN PBRN duration. The program is time-stamped so that the time staff (CK) conducted 135 interviews at CDN CHCs. All required to complete the study is captured. Users can log off questionnaires were administered in English. Participants and log on as often as they wish, allowing the questionnaire to received a $20 honorarium for their time. The mean time be completed in more than one session. To help the person needed to complete the questionnaire was 33 min (range: 12– providing the bleeding history to better understand the questions and give accurate responses, some of the questionsinclude visual aids, for example photographs of petechiae. Data are stored in a secure, Web-accessible MySQL database.
Investigators from other institutions can review all of the All analyses were performed with PASW 18.0 (PASW, components of the system, including the database, at http:// Chicago, IL, USA). The questionnaire includes both top-level screening questions and detailed follow-up questions (e.g.
pertaining to the frequency or duration of bleeding symptoms).
For this study, 36 top-level questions were selected for detailed analysis on the basis of their: (i) similarity to questions reported The study protocol was approved by the institutional review in the existing literature; (ii) reflecting the presence or absence boards of both the Rockefeller University and the Clinical of a bleeding symptom rather than attributes such as the Directors Network (CDN; http://www.CDNetwork.org), a frequency, duration or severity of a symptom; and (iii) eliciting non-profit primary care practice-based research network dichotomous responses that were suitable for binary logistic (PBRN) and clinician training organization that conducts clinical and translational research studies in community Both intra-rater and inter-rater reliabilities were evaluated.
health centers (CHCs). Bleeding symptoms were obtained To assess the consistency of responses to these questions by the from 500 healthy individuals, of whom 365 were seen at same subject over time (intra-rater reliability), 30 individuals Rockefeller University and 135 were seen at two separate were recalled 6–9 months after the questionnaire was initially CDN-member CHCs (75 at Metropolitan Family Health administered, and the questionnaire was administered again.
Network, Jersey City, NJ, and 60 at Newark Community Of the 36 questions analyzed, 36% had complete concordance Health Center, Newark, NJ). Potential participants were between the first and second administration, 36% had 90–99% eligible if they met the following criteria: age ‡ 18 years; self- concordance, and 28% had concordances that ranged from assessment as being generally healthy; and self-assessment as 56% to 89%. To assess whether the person administering the being able to accurately read and answer questions in questionnaire had any influence on the response, one of us English. Exclusion criteria included a diagnosis of any (ACM) reviewed 31 randomly selected audio recordings bleeding disorder; hepatic or renal disease; malignancy conducted by other interviewers, and completed a second requiring treatment within 1 year prior to enrollment; use questionnaire for each participant based on the recording of any medications with known anticoagulant or antiplatelet (inter-rater reliability). The recorded responses were completely properties other than aspirin or non-steroidal anti-inflamma- concordant for 30 of the 36 questions, four questions had tory drugs within 30 days of enrollment; or any other > 90% concordance, and two questions had 82% concor- medical or psychological condition that would impair the dance. An analysis of responses by site of administration participantÕs ability to accurately respond to questions about revealed minor differences in a few symptoms, but the analysis was confounded by the different demographics at the sites and Participants recruited at Rockefeller were identified through the relatively small number of individuals reporting the online and print advertisements seeking healthy volunteers.
symptoms. Thus, no adjustments were made for the site of Participants recruited through CHCs were patients identified in the waiting room during routine primary care visits who were To test whether the prevalence of the 36 bleeding symptoms willing to complete the interview. At Rockefeller, 372 individ- differed according to one or more demographic characteris- uals responded to advertisements and passed telephone tics, a binary multiple logistic regression model was con- screening, of which 365 (98%) were enrolled. At the CHCs, structed for each symptom. For each model, the question Ó 2010 International Society on Thrombosis and Haemostasis Bleeding symptoms in healthy individuals 103 response was the dependent variable, and sex, race, ethnicity, 86 years), and 63% of subjects were female. The racial age and the frequency of aspirin use were the independent distribution was as follows: 37% African-American, 43% variables. For each question, individuals who answered ÔdonÕt Caucasian, 8% Asian, < 1% Pacific Islander, and 4% more rememberÕ were excluded from analysis. For the aspirin than one race; 8% preferred not to report race. The ethnic analysis, ÔfrequentÕ aspirin users were defined as those who distribution was 19% Hispanic and 80% non-Hispanic; 1% used aspirin once a week or more often (10%), ÔinfrequentÕ aspirin users were defined as those who used aspirin less than The frequencies of the 36 bleeding symptoms analyzed are once a week (49%), and ÔneverÕ users were defined as those displayed in Table 3. Symptom frequencies ranged from 0% who denied using aspirin (39%); 1% of respondents did not (teething bleeding, hemorrhagic stroke, circumcision bleeding, remember their frequency of aspirin use. Prior to regression and umbilical cord bleeding) to 47% (heavy menses). It is of analysis, multicollinearity among the independent variables note that 18 of the 36 symptoms were reported by fewer than was tested with the Spearman rank correlation coefficient for 5% of subjects who responded to the question.
ratio and ordinal variables and CramerÕs V for nominalvariables. From this analysis, Hispanic ethnicity was found to be correlated with African-American and Caucasian race.
Therefore, ethnicity was excluded when race was evaluated as The total number of symptoms was not normally distributed an independent variable, and race was excluded when (Fig. 1). When all symptoms were considered, women ethnicity was evaluated as an independent variable. Model reported more bleeding symptoms than men (P < 0.01 by fit was tested using the chi-squared goodness-of-fit and Mann–Whitney U-test; Fig. 1A). Men reported a median of Hosmer–Lemeshow tests [22]. For models with adequate fit, one symptom (interquartile range [IQR] 0–3) and women adjusted odds ratios (ORs) were calculated for independent reported a median of two symptoms (IQR 1–4). However, variables. For each symptom, the proportion of variance after removal of sex-specific bleeding symptoms (heavy explained by the regression model was estimated using the menses, treatment for heavy menses, bleeding during preg- Nagelkerke R2 [22]. Differences with P-values < 0.05 were nancy, bleeding at delivery, postpartum bleeding, and, for defined as statistically significant. No adjustments for multiple men, circumcision bleeding), men and women both reported comparisons were made in these exploratory analyses.
a median of one symptom (IQR 0–3 for both men andwomen, P = 0.50 by Mann–Whitney U-test; Fig. 1B).
When individual symptoms were analyzed together by logistic regression, however, differences by sex were noted Demographic characteristics are displayed in Table 2. The (Table 4). Thus, easy bruising was more common among mean age was 43 years (standard deviation ± 13.8; range 19– women than among men (24% vs. 7%, OR 4.78, 95% Table 2 Demographic characteristics by site of enrollment *For Rockefeller University vs. Community Health Centers.  FisherÕs exact test. àT-test. §Chi-square.
Ó 2010 International Society on Thrombosis and Haemostasis Table 3 Frequencies of selected bleeding symptoms There was a trend towards more bleeding symptoms in Caucasians (median 2, IQR 1–3) vs. either African-Americans (median 1, IQR 0–3) or Asians (median 1, IQR 0–3, P = 0.07 by Kruskal–Wallis test). Epistaxis was less common among African-Americans than among Caucasians (18% vs. 29%, OR 0.55, 95% CI 0.34–0.90, P = 0.02), as was bleeding in oraround the eye, which included conjunctival hemorrhage, retinal hemorrhage, and bleeding behind the eye (3% vs. 10%, OR 0.29, 95% CI 0.11–0.75, P = 0.01).
Hispanics reported a similar median number of symptoms (median 2, IQR 0–3) as non-Hispanics (median 2, IQR 0–3, P = 0.54). When ethnicity was substituted for race as an independent variable, no symptoms were associated with Age was not associated with the total number of symptoms (Spearman r = 0.05, P = 0.24), but the odds of reporting epistaxis decreased with age (OR 0.97 per year, 95% CI 0.96– 0.99, P < 0.01). Increasing age was also associated with a higher reported frequency of dark stools (OR 1.02 per year, 95% CI 1.00–1.04, P = 0.02); however, among the 72 indi- viduals who reported a history of dark stools, 68% had been told that the change in color was attributable to iron therapy, and only 6% had been told that the change in color was caused There was a trend towards an increased median number of symptoms in infrequent aspirin users (median 2, IQR 1–3) vs.
those who never used aspirin (median 1, IQR 0–3) or who used it frequently (median 1, IQR 1–2, P = 0.08 by Kruskal– Wallis test). Easy bruising was more common among infre- quent than among frequent aspirin users (21% vs. 8%, OR 4.18, 95% CI 1.39–12.58, P = 0.01), with those who reported not taking aspirin having an intermediate frequency (15%). A history of heavy menses was also more common among infrequent aspirin users than among frequent aspirin users (56% vs. 38%, OR 3.03, 95% CI 1.25–7.32, P = 0.01), with those who reported not taking aspirin again having an *Exact question wording is available at http://ds9.rockefeller.edu/RUBHPSR/Documentation.html.  Some questions were contingent Contribution of sex, race, age and aspirin use to variability upon previous answers; therefore, not all questions were asked of allrespondents.
For 30 symptoms, no significant associations with sex, race, age confidence interval [CI] 2.50–9.16, P < 0.01), as was veni- or aspirin use were identified. The contribution of sex, race, age puncture bruising (10% vs. 3%, OR 3.04, 95% CI 1.22–7.59, and aspirin use to the remaining six symptoms, as defined by Nagelkerke R2, are summarized in Table 4. For instance, the Ó 2010 International Society on Thrombosis and Haemostasis Bleeding symptoms in healthy individuals 105 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Number of symptoms, excluding sex-specific symptoms Fig. 1. Number of bleeding symptoms in men and women. (A) Men reported fewer bleeding symptoms than women (P < 0.01 by Mann–Whitney U-test). (B) When heavy menses, treatment for heavy menses, bleeding during pregnancy, bleeding at delivery, postpartum bleeding and circumcision bleedingwere excluded, men and women reported similar frequencies of bleeding symptoms (P = 0.50 by Mann–Whitney U-test).
Table 4 Results of logistic regression analyses of questions with statistically significant differences in responses by one or more demographic characteristics *Odds ratio per additional year of age.  Odds ratio vs. Caucasian race. àOdds ratio vs. male sex. §Defined as aspirin use less than once per week;odds ratio vs. aspirin use more than once per week.
combined R2 for age and race with regard to epistaxis was 0.08; established whether applying adjustments for these character- that is, the two variables accounted for 8% of the observed istics can improve the diagnostic value of the bleeding history.
variability in epistaxis. Sex, race, age and aspirin use accounted The frequencies of most bleeding symptoms in our popula- for only 6–13% of the variability in reported symptoms.
tion were within the ranges previously reported by otherinvestigators (Table 1). For instance, 25% of our respondentsreported epistaxis, as compared with a weighted average of 23% in previous studies, and 18% of our respondents reported We used a novel questionnaire and Web-based system to easy bruising, as compared with a weighted average of 20% in collect comprehensive bleeding histories from 500 healthy previous studies. Of the 13 symptoms reported in Table 1, nine individuals. We found that 26% of subjects reported no symptoms were within 5% of the reported weighted average.
symptoms, and that 18 of the 36 symptoms were reported by Greater differences were observed for menorrhagia (47% in fewer than 5% of the subjects. The frequencies of epistaxis, our study vs. 35% in the literature), tooth extraction bleeding easy bruising, bruising after venipuncture, heavy menses, dark (18% vs. 6%), gum bleeding (4% vs. 28%), and hematochezia stools and ophthalmic bleeding exhibited variations according (19% vs. 7%). It is likely that some of the variation is to sex, racial background, age, and frequency of aspirin use, attributable to differences in question wording. For instance, but these characteristics accounted for only an estimated 6– the lower frequency of gum bleeding that we found probably 13% of the total variability in the reported frequencies of these reflects the requirement in our questionnaire, but not in those six symptoms. Our findings suggest that the definition of of other authors, that the gum bleeding last at least 5 min.
ÔnormalÕ bleeding that is not sex-specific varies relatively little Differences in the populations studied may also contribute to by sex, race, age, or aspirin use; thus, it remains to be Ó 2010 International Society on Thrombosis and Haemostasis Although women reported a higher median number of easy bruising and heavy menses were more common among symptoms than men when viewed from the standpoint of all infrequent aspirin users than among frequent aspirin users, bleeding symptoms, this difference did not persist after with non-users having intermediate values. We have no simple exclusion of sex-specific questions (e.g. menstrual bleeding explanation for this finding, but frequent aspirin use was and postpartum bleeding). Thus, bleeding scores that are based reported by only 10% of responders, and it is possible that on the number of bleeding symptoms need to be adjusted for individuals with heavy menses or easy bruising who were sex, as is done, for example, with the Vicenza bleeding score for frequent aspirin users had been counseled to reduce their VWD [4]. Easy bruising and bruising after venipuncture were, aspirin intake. Unfortunately, the aspirin primary prevention however, more common in women than in men, a finding that studies noted above did not report the frequency of heavy is consistent with the study by Wahlberg et al. [13], in which women reported a higher frequency of spontaneous bruising There is an apparent contradiction between the finding that than men. It is also consistent with studies of primary aspirin men and women reported similar total numbers of non-sex- prophylaxis, where the reported frequencies of excessive specific symptoms and the finding that women reported higher bruising among the control groups not taking aspirin were frequencies of easy bruising and venipuncture bruising. The 43% among women [23] and 9–13% among men [24,25].
reason why the increased frequency of bruising in women did The literature on differences in bleeding symptoms as a not affect the median value for all symptoms was that it was function of race and ethnicity is sparse. In the only study counterbalanced by statistically insignificant increases in identified that addressed these issues, the authors found a lower symptom frequencies among men in several categories (gum frequency of epistaxis among Asians than among Caucasians bleeding, lip bleeding, tooth extraction bleeding, hemoptysis, [18], a finding that we did not observe. We found that epistaxis minor cut bleeding, hematochezia, melena, eye bleeding, and ophthalmic bleeding were reported with lower frequency surgical bleeding, and trauma bleeding). Similarly, although by African-Americans than by Caucasians. It is possible that we observed differences in several symptoms by race or this difference can be explained by the higher von Willebrand frequency of aspirin use, the total number of symptoms did not factor levels reported in African-Americans than in Caucasians vary by race, age, or frequency of aspirin use, despite [26–28], as epistaxis is a cardinal symptom of VWD [1,4].
statistically insignificant variations within individual categories.
However, the African-American women in our population did The Vicenza group has demonstrated the diagnostic and not report a lower frequency of heavy menses, another prognostic utility of their questionnaire for VWD [4–6]. Based common manifestation of VWD [4,29], than Caucasian on their experience and expertise, they excluded from consid- eration all bleeding symptoms that they defined as Ôtrivial,Õ and As our bleeding history questionnaire is based on the devised their bleeding score on the basis of the severity of cumulative prevalence of a list of symptoms, we considered the individual symptoms. Our questionnaire, in contrast, includes possibility that older individuals would report more symptoms questions on both major and minor bleeding symptoms. This because they had more time in which to experience symptoms.
explains why we found a lower percentage of healthy individ- Although a correlation between increasing age and higher uals who did not report any bleeding symptoms (26%) than did bleeding scores has been reported in individuals with bleeding the Vicenza group (77%) [4]. We are currently employing our disorders, previous reports have not found this correlation in questionnaire to identify the symptoms of patients with mild healthy individuals [4,5]. We also found few associations bleeding disorders, and will compare these data with the data between older age and more bleeding symptoms in our healthy that we have obtained in the healthy adult cohort described in adult population. Paradoxically, we found a modest decrease in this study, using several different statistical methods.
the reported frequency of epistaxis with age, suggesting either Because the reliability of medical histories depends on the that epistaxis is becoming more common or that older ability of subjects to recall symptoms over their entire lifetimes individuals are less likely to remember episodes of epistaxis and to report the data consistently, these studies will also from their youth. We also found an increase in the reported evaluate measures of questionnaire validation such as test– frequency of dark stools with age. However, among the retest and inter-rater reliability. In the current study, we chose individuals who reported dark stools, the symptom could be to use a comprehensive questionnaire, despite its length, classified with reasonable certainty as melena in only 6% of because we did not want to prejudge which questions would cases. This highlights the limited utility of just inquiring about have the best diagnostic and prognostic value. Going forward, dark stools, and thus the importance of following up screening we will compare the responses of the healthy adults in this questions with questions that provide more details.
study with those of patients with bleeding disorders, so as to Although aspirin use has been reported to increase the identify those questions that are most reliable and/or of the frequency of a number of bleeding symptoms, including easy greatest diagnostic value. We will then create a shorter bruising (which ranged from 9% to 43% in controls and from questionnaire that focuses on obtaining the most valuable 14% to 53% in studies on aspirin in primary prevention studies data. The potential advantages and drawbacks of our approach [23–25]), we did not exclude individuals taking aspirin, because we wanted a sample that represented as closely as possible the The long-term goal of our research is to standardize the demographics of our community. Unexpectedly, we found that collection of bleeding histories by developing Web-based Ó 2010 International Society on Thrombosis and Haemostasis Bleeding symptoms in healthy individuals 107 instruments that can be used by investigators across different boom J, Schneppenheim R, Budde U, Ingerslev J, Vorlova Z, Habart sites and studies. This has the potential to aggregate large D, Holmberg L, Lethagen S, Pasi J, Hill F, et al. A quantitativeanalysis of bleeding symptoms in type 1 von Willebrand disease: amounts of de-identified phenotypic data, so as to increase the results from a multicenter European study (MCMDM-1 VWD).
power to detect scientifically and medically important correla- tions with genetic and environmental data. As a first step in this 6 Tosetto A, Castaman G, Rodeghiero F. Evidence-based diagnosis of process, an electronic version of our questionnaire is available type 1 von Willebrand disease: a Bayes theorem approach. Blood 2008; to investigators at https://ds9.rockefeller.edu/RUBHPSR/.
7 Federici AB, Mannucci PM, Castaman G, Baronciani L, Bucciarelli P, Canciani MT, Pecci A, Lenting PJ, De Groot PG. Clinical and molecular predictors of thrombocytopenia and risk of bleeding inpatients with von Willebrand disease type 2B: a cohort study of 67 We would like to thank D. Rubin and M. Musen of Stanford patients. Blood 2009; 113: 526–34.
University and S. Mollah of Rockefeller University for their 8 Bowman M, Mundell G, Grabell J, Hopman WM, Rapson D, Lillicrap D, James P. Generation and validation of the Condensed valuable conceptual and practical contributions to the design of MCMDM-1VWD Bleeding Questionnaire for von Willebrand disease.
the Bleeding History Phenotyping System, D. Brassil and D.
Bernal-Messinger for their assistance in administering the 9 Bowman M, Riddel J, Rand ML, Tosetto A, Silva M, James PD.
Bleeding History Questionnaire, and the staff and patients of Evaluation of the diagnostic utility for von Willebrand disease of a the Metropolitan Family Health Center Network (P. Beaty, pediatric bleeding questionnaire. J Thromb Haemost 2009; 7:1418–21.
10 Biss TT, Blanchette VS, Clark DS, Bowman M, Wakefield CD, Silva M, Chief Medical Officer) and the Newark Community Health Lillicrap D, James PD, Rand ML. Quantitation of bleeding symptoms Centers, Inc. (N. Tham, Chief Medical Officer) for their in children with von Willebrand disease: use of a standardized pediatric bleeding questionnaire. J Thromb Haemost 2010; 8: 950–6.
11 Plug I, Mauser-Bunschoten EP, Brocker-Vriends AH, van Amstel HK, van der Bom JG, van Diemen-Homan JE, Willemse J, Rosendaal FR. Bleeding in carriers of hemophilia. Blood 2006; 108: 52–6.
12 Dilley A, Drews C, Miller C, Lally C, Austin H, Ramaswamy D, In accord with federal law and the policies of the Research Lurye D, Evatt B. von Willebrand disease and other inherited bleeding Foundation of the State University of New York, B. S. Coller disorders in women with diagnosed menorrhagia. Obstet Gynecol 2001; has a royalty interest in abciximab (Centocor), and in accord with federal law and the policies of the Mount Sinai School of 13 Wahlberg T, Blomback M, Hall P, Axelsson G. Application of indi- Medicine, B. S. Coller has a royalty interest in the VerifyNow cators, predictors and diagnostic indices in coagulation disorders. I.
Evaluation of a self-administered questionnaire with binary questions.
assay system (Accumetrics). In addition, he is a consultant to Methods Inf Med 1980; 19: 194–200.
Accumetrics and is an inventor of an aIIbb3 antagonist 14 Sramek A, Eikenboom JC, Briet E, Vandenbroucke JP, Rosendaal compound identified by high-throughput screening. This study FR. Usefulness of patient interview in bleeding disorders. Arch Intern was supported by grants KL2RR024142 and UL1RR024143 from the National Center for Research Resources (NCRR), a 15 Nosek-Cenkowska B, Cheang MS, Pizzi NJ, Israels ED, Gerrard JM.
Bleeding/bruising symptomatology in children with and without component of the National Institutes of Health (NIH), and the bleeding disorders. Thromb Haemost 1991; 65: 237–41.
NIH Roadmap for Medical Research, as well as funds from 16 Quiroga T, Goycoolea M, Panes O, Aranda E, Martinez C, Belmont S, Stony Brook University. The contents are solely the respon- Munoz B, Zuniga P, Pereira J, Mezzano D. High prevalence of sibility of the authors and do not necessarily represent the bleeders of unknown cause among patients with inherited mucocuta- neous bleeding. a prospective study of 280 patients and 299 controls.
Haematologica 2007; 92: 357–65.
17 McKay H, Derome F, Haq MA, Whittaker S, Arnold E, Adam F, Heddle NM, Rivard GE, Hayward CP. Bleeding risks associatedwith inheritance of the Quebec platelet disorder. Blood 2004; 104: 159– 1 Coller BS, Schneiderman PI. Clinical evaluation of hemorrhagic dis- orders: the bleeding history and differential diagnosis of purpura. In: 18 Daniel M, Jaberoo MC, Stead RE, Reddy VM, Moir AA. Is admis- Hoffman R, Benz EJ, Shattil SJ, Furie B, Silberstein LE, McGlave P, sion for epistaxis more common in Caucasian than in Asian people? A eds. Hematology: Basic Principles and Practice, 5th edn. New York: preliminary study. Clin Otolaryngol 2006; 31: 386–9.
Churchill Livingstone, 2004: 1851–76.
19 National Institutes of Health. NIH policy on reporting race and 2 Chee YL, Crawford JC, Watson HG, Greaves M. Guidelines on the ethnicity data: subjects in clinical research. http://grants.nih.gov/ assessment of bleeding risk prior to surgery or invasive procedures.
grants/guide/notice-files/not-od-01-053.html. Accessed 13 September British Committee for Standards in Haematology. Br J Haematol 20 Mauer AC, Barbour EM, Khazanov NA, Levenkova N, Mollah SA, 3 Eckman MH, Erban JK, Singh SK, Kao GS. Screening for the risk for Coller BS. Creating an ontology-based human phenotyping system: bleeding or thrombosis. Ann Intern Med 2003; 138: W15–24.
the Rockefeller University bleeding history experience. Clin Transl Sci 4 Rodeghiero F, Castaman G, Tosetto A, Batlle J, Baudo F, Cappelletti A, Casana P, De BN, Eikenboom JC, Federici AB, Lethagen S, Linari 21 Bodenreider O. The Unified Medical Language System (UMLS): S, Srivastava A. The discriminant power of bleeding history for the integrating biomedical terminology. Nucleic Acids Res 2004; 32: D267– diagnosis of type 1 von Willebrand disease: an international, multicenter study. J Thromb Haemost 2005; 3: 2619–26.
22 Tabachnick BG, Fidell LS. Logistic regression. In: Tabachnick BG, 5 Tosetto A, Rodeghiero F, Castaman G, Goodeve A, Federici AB, Fidell LS, eds. Using Multivariate Statistics, 5th edn. New York: Allyn Batlle J, Meyer D, Fressinaud E, Mazurier C, Goudemand J, Eiken- Ó 2010 International Society on Thrombosis and Haemostasis 23 Ridker PM, Cook NR, Lee IM, Gordon D, Gaziano JM, Manson JE, age, race, sex, and risk factors for atherosclerosis. The Atherosclerosis Hennekens CH, Buring JE. A randomized trial of low-dose aspirin in Risk in Communities (ARIC) Study. Thromb Haemost 1993; 70: 380– the primary prevention of cardiovascular disease in women. N Engl J 29 Kadir RA, Economides DL, Sabin CA, Owens D, Lee CA. Frequency 24 Steering Committee of the PhysiciansÕ Health Study Research Group.
of inherited bleeding disorders in women with menorrhagia. Lancet Final report on the aspirin component of the ongoing PhysiciansÕ Health Study. N Engl J Med 1989; 321: 129–35.
30 Wahlberg TB. A method for the evaluation of clinical information, 25 The Medical Research Council’s General Practice Research Frame- exemplified for bleeding symptoms in non-severe von WillebrandÕs work. Thrombosis prevention trial: randomised trial of low-intensity disease type I. Methods Inf Med 1984; 23: 143–6.
oral anticoagulation with warfarin and low-dose aspirin in the primary 31 Mauser Bunschoten EP, van Houwelingen JC, Sjamsoedin Visser EJ, prevention of ischaemic heart disease in men at increased risk. Lancet van Dijken PJ, Kok AJ, Sixma JJ. Bleeding symptoms in carriers of hemophilia A and B. Thromb Haemost 1988; 59: 349–52.
26 Miller CH, Haff E, Platt SJ, Rawlins P, Drews CD, Dilley AB, Evatt B.
32 Drews CD, Dilley AB, Lally C, Beckman MG, Evatt B. Screening Measurement of von Willebrand factor activity: relative effects of ABO questions to identify women with von Willebrand disease. J Am Med blood type and race. J Thromb Haemost 2003; 1: 2191–7.
27 Green D, Ruth KJ, Folsom AR, Liu K. Hemostatic factors in the 33 Friberg B, Orno AK, Lindgren A, Lethagen S. Bleeding disorders Coronary Artery Risk Development in Young Adults (CARDIA) among young women: a population-based prevalence study. Acta Study. Arterioscler Thromb 1994; 14: 686–93.
Obstet Gynecol Scand 2006; 85: 200–6.
28 Conlan MG, Folsom AR, Finch A, Davis CE, Sorlie P, Marcucci G, Wu KK. Associations of factor VIII and von Willebrand factor with Ó 2010 International Society on Thrombosis and Haemostasis

Source: http://bioontology.stanford.edu/sites/default/files/impact%20of%20sex,%20age,%20race.pdf