Internet as a medium to seek partners among men who have sex with men

While there are reports of developing sexual relationships on the Internet (I) among MSM, there are few reports that have examined the process of developing sexual relationships on the I and comparing to that in real life (IRL). This study examines the process to provide insight into how MSM make decisions about courtship, engages in negotiations for sex, and choose sexual partners and examines the comparative sexual risks taken between I vs. IRL negotiation. This self-selected convenience sample at a national level (n=1001) of MSM recruited through the I, systematically explored the different steps, the process of courtship in a flow chart of I and IRL dating to portray the process of filtering, courtship and/or negotiation for sex. Risk behaviors in both environments are presented along with interactions that create predictable sequences or "scripts". These sequences constitute 'filtering' and 'sexual positioning'. Differences between I & IRL suggest discussion of HIV/STD status to have consistent differences for all variables except 'unprotected sex' meaning no condom use. There was more communication on the I in regards to self revealing information or variables relating to reducing risks which enable 'filtering' (including serosorting). Data indicate more steps in the I process, providing more complex, multiple steps to filter and position with regard not only to HIV/STD risk but also to negotiate position for complementary sexual interest. The study established a pattern of MSM's courtships or negotiation for sex and a pattern of acquisition, and more I negotiation. Data suggest negotiation opportunities which could lend to intervention to advise people how to negotiate safely. ^ Previous studies have reviewed MSM and drug use. This is a study to review the process of drug use associated with sexual behavior regarding the Internet (I) and in real life (IRL) using a self-selected, convenience sample of MSM (n=1001) recruited nation-wide through the Internet. Data on MSM and drugs illustrate the Internet being used as a tool to filter for drug use among MSM. MSM's drug use in both environments highlights the use of sexual performance drugs with an IRL pursuit of intimacy or negotiation for sex. IRL encounters were more likely to involve drug use (both recreational and sexual performance-enhancing) than Internet encounters. This may be due to more IRL meetings occurring at bars, clubs or parties where drug use is a norm. Compared with IRL, the Internet may provide a venue for persons who do not want to use drugs to select partners with similar attitudes. This suggests that filtering may be occurring as part of the internet negotiation. Data indicated that IRL persons get drunk/high before having sex in past 60 days significantly more often than Internet participants. Age did not alter the pattern of results. Thus drug filtering is really not recreational drug filtering or selecting for PNP, but appears to be situationally-based. Thus, it should perhaps be seen as another form of filtering to select drug-free partners, rather than using the Internet to specifically recruit and interact with other recreational drug users. ^

Using citation data to improve retrieval from MEDLINE.

OBJECTIVE: To determine whether algorithms developed for the World Wide Web can be applied to the biomedical literature in order to identify articles that are important as well as relevant. DESIGN AND MEASUREMENTS A direct comparison of eight algorithms: simple PubMed queries, clinical queries (sensitive and specific versions), vector cosine comparison, citation count, journal impact factor, PageRank, and machine learning based on polynomial support vector machines. The objective was to prioritize important articles, defined as being included in a pre-existing bibliography of important literature in surgical oncology. RESULTS Citation-based algorithms were more effective than noncitation-based algorithms at identifying important articles. The most effective strategies were simple citation count and PageRank, which on average identified over six important articles in the first 100 results compared to 0.85 for the best noncitation-based algorithm (p < 0.001). The authors saw similar differences between citation-based and noncitation-based algorithms at 10, 20, 50, 200, 500, and 1,000 results (p < 0.001). Citation lag affects performance of PageRank more than simple citation count. However, in spite of citation lag, citation-based algorithms remain more effective than noncitation-based algorithms. CONCLUSION Algorithms that have proved successful on the World Wide Web can be applied to biomedical information retrieval. Citation-based algorithms can help identify important articles within large sets of relevant results. Further studies are needed to determine whether citation-based algorithms can effectively meet actual user information needs.

Internet use, exposure to Internet pornography, and sexual behavior among middle school youth

The purpose of this study is to examine associations between adolescents’ Internet use, exposure to pornography online, and sexual behavior. This cross-sectional study examines data collected from an HIV, sexually transmitted infection, and pregnancy prevention program being evaluated in inner-city middle schools. Chi-squares were used to examine differences in Internet use and exposure to Internet pornography by gender, race/ethnicity, and sexual behavior. Univariate and multivariate logistic regression were used to examine associations between Internet use, exposure to Internet pornography, and sexual behavior. Ninety-four percent of students have used the Internet. Sixty-two percent of students had accidentally seen pornography on the Internet and 35% had purposefully viewed pornography online. Students who experienced sexual solicitation and who purposefully viewed pornography online were more likely to report lifetime and current sexual behavior. Additional research is needed to understand the impact of Internet use and exposure to Internet pornography on adolescents’ sexual behavior.^

Trends in health information seeking among cancer and non-cancer adults between 2003 and 2005: A descriptive analysis of Health Information National Trends Survey (HINTS) data

Cancer is the second leading cause of death in the United States. With the advent of new technologies, changes in health care delivery, and multiplicity of provider types that patients must see, cancer care management has become increasingly complex. The availability of cancer health information has been shown to help cancer patients cope with the management and effects of their cancers. As a result, more cancer patients are using the internet to find resources that can aid in decision-making and recovery. ^ The Health Information National Trends Survey (HINTS) is a nationally representative survey designed to collect information about the experiences of cancer and non-cancer adults with health information sources. The HINTS survey focused on both conventional sources as well as newer technologies, particularly the internet. This study is a descriptive analysis of the HINTS 2003 and HINTS 2005 survey data. The purpose of the research is to explore the general trends in health information seeking and use by US adults, and especially by cancer patients. ^ From 2003 to 2005, internet use for various health-related activities appears to have increased among adults with and without cancer. Differences were found between the groups in the general trust in information media, particularly the internet. Non-cancer respondents tended to have greater trust in information media than cancer respondents. ^ The latter portion of this work examined characteristics of HINTS respondents that were thought to be relevant to how much trust individuals placed in the internet as a source of health information. Trust in health information from the internet was significantly greater among younger adults, higher-earning households, internet users, online seekers of health or cancer information, and those who found online cancer information useful. ^

America's state health departments and the enhancement of health communication through podcasting

Perhaps it was foreshadowing the influence emerging technologies would have on health when the term "podcast" beat out "bird flu" for the 2005 word of the year, an honor given by The New Oxford American Dictionary. From medical school courses to medical journal summaries, podcasting has found a niche in the health field and with studies showing a high proportion of people using the Internet to seek health information, it is imperative that the online information be accurate and easily accessible. With the responsibility of health departments to reach out to their respective communities with effective health communication strategies, this study assessed the proportion of 50 states' and the District of Columbia's health departments utilizing podcasting as a tool for health communication. Additionally, to assess any trends, the prevalence of podcasting was compared to select state demographic characteristics (age, sex, and median income), the organization of the health department (freestanding or super-agency), and the respective United Health Foundation 2007 health ranking. ^ Prevalence data were collected from each state health department's website to find evidence of podcasting to any extent. If a podcast was present, characteristics including creator, release frequency, and transcript were further assessed. The study found that 51% (26/51) of all health departments were utilizing podcasts in some capacity and almost 20% (5/26) of these had created their own podcasts. The most common use of podcasting was to link to outside podcast resources, most notably, the Centers for Disease Control's podcast series. No significant associations were found between the state-specific variables and the podcasting outcomes; however, higher percentages of young adults in some states suggest potential podcasting opportunities for targeting these known podcast users with age-specific health messages. Another recommendation is a future assessment of local health departments' use of podcasting as their smaller, more defined target audiences may be a more efficient use of podcasting as a health communication tool. Additionally, there is a need for evaluations of podcasts' overall effectiveness as a health communication tool to (1) reach a target audience; and (2) convey a specific health message. In conclusion, the findings from this project illuminate the extent of podcast influence in states' and the District of Columbia's health departments as a health communication tool; however evaluations of effectiveness are imperative for future studies.^

A qualitative assessment of the impact of Texas policy changes on nutrition in child care centers in Travis County, Texas

Background. Over half of children in the United States under age five spend 32 hours a week in child care, facilities, where they consume approximately 33-50% of their food intake. ^ Objectives. The aim of this research was to identify the effects of state nutrition policies on provision of food in child care centers. ^ Subjects. Eleven directors or their designee from ten randomly selected licensed child care centers in Travis County, Texas were interviewed. Centers included both nonprofit and for-profit centers, with enrollments ranging from 19 to 82. ^ Methods. Centers were selected using a web-based list of licensed child care providers in the Austin area. One-on-one interviews were conducted in person with center directors using a standard set of questions developed from previous pilot work. Interview items included demographic data, questions about state policies regarding provision of foods in centers, effects of policies on child care center budgets and foods offered, and changes in the provision of food. All interviews were audiotaped and transcribed, and themes were identified using standard qualitative techniques. ^ Results. Four of the centers provided both meals and snacks, four provided snacks only, and two did not provide any food. Directors of centers that provided food were more likely to report adherence to the Minimum Standards than directors of centers that did not. In general, center directors reported that the regulations were loosely enforced. In contrast, center directors were more concerned about a local city-county regulation that required food permits and new standards for kitchens. Most of these local regulations were cost prohibitive and, as a result, centers had changed the types of foods provided, which included providing less fresh produce and more prepackaged items. Although implementation of local regulations had reduced provision of fruits and vegetables to children, no adjustments were reported for allocation of resources, tuition costs or care of the children. ^ Conclusions. Qualitative data from a small sample of child care directors indicate that the implementation and accountability of food- and nutrition-related guidelines for centers is sporadic, uncoordinated, and can have unforeseen effects on the provision of food. A quantitative survey and dietary assessment methods should be conducted to verify these findings in a larger and more representative sample.^

Hepatitis C and human immunodeficiency virus infections in injecting drug users in drug treatment centers in Vietnam

Background. Injecting drug users (IDUs) are at risk of infection with Hepatitis C Virus (HCV) and Human Immunodeficiency Virus (HIV). Independently, each of these viruses is a serious threat to health, with HIV ravaging the body’s immune system, and HCV causing cirrhosis, liver cancer and liver failure. Co-infection with HIV/HCV weakens the response to antiretroviral therapy in HIV patients. IDUs with HIV/HCV co-infection are at a 20 times higher risk of having liver-related morbidity and mortality than IDUs with HIV alone. In Vietnam, studies to ascertain the prevalence of HIV have found high rates, but little is known about their HCV status. ^ Purpose. To measure the prevalence of HCV and HIV infection and identify factors associated with these viruses among IDUs at drug treatment centers in northern Vietnam. ^ Methods. A cross-sectional study was conducted from November 2007 to February 2008 with 455 injecting drug users aged 18 to 39 years, admitted no more than two months earlier to one of four treatment centers in Northern Vietnam (Hatay Province) (response rate=95%). Participants, all of whom had completed detoxification and provided informed consent, completed a risk assessment questionnaire and had their blood drawn to test for the presence of antibody-HCV and antibody-HIV with enzyme immuno assays. Univariate and multivariable logistic regression models were utilized to explore the strength of association using HIV, HCV infections and HIV/HCV co-infection as outcomes and demographic characteristics, drug use and sexual behaviors as factors associated with these outcomes. Unadjusted and adjusted odds ratios and 95% confidence intervals were calculated. ^ Results. Among all IDU study participants, the prevalence of HCV alone was 76.9%, HIV alone was 19.8%. The prevalence of HIV/HCV co-infection was 92.2% of HIV-positive and 23.7% of HCV-positive respondents. No sexual risk behaviors for lifetime, six months or 30 days prior to admission were significantly associated with HCV or HIV infection among these IDUs. Only duration of injection drug use was independently associated with HCV and HIV infection, respectively. Longer duration was associated with higher prevalence. Nevertheless, while HCV infection among IDUs who reported being in their first year of injecting drugs were lower than longer time injectors, their rates were still substantial, 67.5%. ^ Compared with either HCV mono-infection or HIV/HCV non-infection, HIV/HCV co-infection was associated with the length of drug injection history but was not associated with sexual behaviors. Higher education was associated with a lower prevalence of HIV/HCV co-infection. When compared with HIV/HCV non-infection, current marriage was associated with a lower prevalence of HIV/HCV co-infection. ^ Conclusions. HCV was prevalent among IDUs from 18 to 39 years old at four drug treatment centers in northern Vietnam. Co-infection with HCV was predominant among HIV-positive IDUs. HCV and HIV co-infection were closely associated with the length of injection drug history. Further research regarding HCV/HIV co-infection should include non-injecting drug users to assess the magnitude of sexual risk behaviors on HIV and HCV infection. (At these treatment centers non-IDUs constituted 10-20% of the population.) High prevalence of HCV prevalence among IDUs, especially among HIV-infected IDUs, suggests that drug treatment centers serving IDUs should include not only HIV prevention education but they should also include the prevention of viral hepatitis. In addition, IDUs who are HIV-positive need to be tested for HCV to receive the best course of therapy and achieve the best response to HIV treatment. These data also suggest that because many IDUs get infected with HCV in the first year of their injection drug career, and because they also engaged in high risk sexual behaviors, outreach programs should focus on harm reduction, safer drug use and sexual practices to prevent infection among drug users who have not yet begun injecting drugs and to prevent further spread of HCV, HIV and co-infection. ^

The association of internet use and depression among spinal cord injury population

The purpose of this study was to exam the relationship between internet use and depression among a population of individuals who have sustained spinal cord injury. This was cross-sectional survey design conducted among spinal cord injury (SCI) patients in the Model Spinal Cord Injury System. We included a total of 1,011 SCI-patients who were interviewed face-to-face or by telephone interview over approximately a three year time period (2004–2006). All data were collected through a telephone survey which included the Patient Health Questionnaire-9 (PHQ-9) to assess depression. We examined various scales of this survey, included a reduced 3-item scale (items 1, 2 and 6) to avoid the presence of somatic symptoms among SCI patients from influencing classification of depression. The frequency of internet usage was grouped as daily/weekly/monthly/non user. Covariates examined as possible confounders included demographic characteristics, occupational status, educational level, injury type, daily function of living, pain level, self-perceived health status and satisfaction with life. We observed a negative association between the frequency of internet use and the level of depression. Daily use of internet was associated with lower PHQ-9 score and depression; however this association did not reach statistical significance after for the mentioned covariates. In conclusion, the factors related to lower depression in SCI patients who use the internet are complicated. Daily internet usage was associated with lower levels of depression. The accuracy of 3-item scale needs further validation and investigation. Further study of internet usage pattern in SCI patient is recommended. ^

Systematic literature review on internet-based interventions to prevent HIV transmission

Objective. To perform a systematic review of the internet based HIV prevention interventions.^ Data Source. Data resources including Ovid Medline, CINAHL (EBSCOhost and Psycinfo) were searched for the relevant articles. Study Selection: Studies were selected if the entire article was in English, Randomized control trial performed after 1998 till current, internet based aimed towards HIV/AIDS prevention. ^ Data Extraction. Each relevant article was coded using a code sheet which consisted of background information, study characteristics, study population information, program details, intervention and results. Data Synthesis: Total of seven relevant articles were coded and the information obtained was incorporated into the evidence tables for comparison.^ Conclusion. Overall the internet based HIV prevention studies have been very successful in recruiting large number of people but has not been able to show effective results due to high attrition. More research needs to be done in the area.^

Internet sex among non-gay identifying men who have sex with men

The aim of this study was to examine the role of the Internet in Internal Homonegativity (IH) among Non gay identifying men who have sex with men (NGI-MSM). This study at University of Texas School of Public Health (UTSPH) had a mixed method research design and consisted of men 18 years of age and older who were residents of the US and Canada. The data were collected using an online survey called 'Men's Sexual Health Survey' which was developed in collaboration with Boston University School of Public Health and Denver Public Health. These surveys were administered in English, which took 30-minutes to complete, and were placed on gay oriented websites and chat rooms. 141 participants were presented with the module relating to IH. A Principal Component Analysis with varimax rotation on the nine questions that asked the participants about their feelings about gay men produced three factors of IH identified as (1) public identification as gay; (2) perception of stigma associated with being gay; and (3) social comfort with gay men. The factors significantly correlated with age, grade completed in school, income, openness about being gay and socializing with gay people, meeting partners online, dating on the Internet, attitude toward condom usage, alcohol and drug use before sex and having unprotected sex with Internet partners. These findings point toward the role of the Internet in determining IH and sexual behavior. Despite the risks, the Internet's popularity and outreach in NGI-MSM makes it an effective medium to spread public health programs.^

Ethnic differences in food availability at home among 3--5 year old preschoolers attending Head Start centers in 2008

Background. The childhood obesity epidemic has disproportionately impacted the lives of low-income, minority preschoolers and their families. Research shows that parents play a major role as "gatekeepers" who control what food is brought into the home and as role models for dietary practices. Currently, there is limited research regarding ethnic differences in families of low-income preschoolers. ^ Objective. The objective of this study was to look at ethnic differences in food availability at home among the low-income families of Hispanic and African American preschoolers attending Head Start centers in Harris County, Texas. ^ Design/Subjects. Descriptive data on food availability at home between Hispanic and African American families were used and analyzed for this study. Parents or primary caregivers (n = 718) of children enrolled at Head Start Centers in Houston, Texas completed the Healthy Home Survey. ^ Methods. In the Healthy Home Survey, participants were asked to answer open-ended questions regarding various types of foods currently available at home, such as fresh, canned or jarred, dried and frozen fruits; fresh, canned or jarred, and frozen vegetables; salty snacks, sweet snacks, candy, and soda. Descriptive analyses were conducted to identify significant differences between Hispanics and African Americans via a paired t-test to compare the means of variables between the study groups and a Pearson's chi-square or Fischer's exact (if cell size was <5) test calculated for food availability (food types) between ethnicities to determine differences in distributions. ^ Results. Although both Hispanics and African Americans reported having all categories of food types at home, there were statistically significant differences between ethnic groups. Hispanics were more likely to have fresh fruits and vegetables at home than African Americans. At the same time, more African American families reported having canned or jarred fruits and canned green/leafy vegetables than Hispanics. More Hispanic families reported having diet, regular, and both diet and regular sodas available compared to African American families. However, high percentages of unhealthy foods (including snacks and candy) were reported by both ethnicities. ^ Conclusions. The findings presented in this study indicate the implicit ethnic differences that exist in the food availability among low-income families of Hispanic and African American preschoolers. Future research should investigate the associations between food availability and children's weight status by ethnicity to identify additional differences that may exist.^

Data Accuracy in Medical Record Abstraction

Clinical Research Data Quality Literature Review and Pooled Analysis We present a literature review and secondary analysis of data accuracy in clinical research and related secondary data uses. A total of 93 papers meeting our inclusion criteria were categorized according to the data processing methods. Quantitative data accuracy information was abstracted from the articles and pooled. Our analysis demonstrates that the accuracy associated with data processing methods varies widely, with error rates ranging from 2 errors per 10,000 files to 5019 errors per 10,000 fields. Medical record abstraction was associated with the highest error rates (70–5019 errors per 10,000 fields). Data entered and processed at healthcare facilities had comparable error rates to data processed at central data processing centers. Error rates for data processed with single entry in the presence of on-screen checks were comparable to double entered data. While data processing and cleaning methods may explain a significant amount of the variability in data accuracy, additional factors not resolvable here likely exist. Defining Data Quality for Clinical Research: A Concept Analysis Despite notable previous attempts by experts to define data quality, the concept remains ambiguous and subject to the vagaries of natural language. This current lack of clarity continues to hamper research related to data quality issues. We present a formal concept analysis of data quality, which builds on and synthesizes previously published work. We further posit that discipline-level specificity may be required to achieve the desired definitional clarity. To this end, we combine work from the clinical research domain with findings from the general data quality literature to produce a discipline-specific definition and operationalization for data quality in clinical research. While the results are helpful to clinical research, the methodology of concept analysis may be useful in other fields to clarify data quality attributes and to achieve operational definitions. Medical Record Abstractor’s Perceptions of Factors Impacting the Accuracy of Abstracted Data Medical record abstraction (MRA) is known to be a significant source of data errors in secondary data uses. Factors impacting the accuracy of abstracted data are not reported consistently in the literature. Two Delphi processes were conducted with experienced medical record abstractors to assess abstractor’s perceptions about the factors. The Delphi process identified 9 factors that were not found in the literature, and differed with the literature by 5 factors in the top 25%. The Delphi results refuted seven factors reported in the literature as impacting the quality of abstracted data. The results provide insight into and indicate content validity of a significant number of the factors reported in the literature. Further, the results indicate general consistency between the perceptions of clinical research medical record abstractors and registry and quality improvement abstractors. Distributed Cognition Artifacts on Clinical Research Data Collection Forms Medical record abstraction, a primary mode of data collection in secondary data use, is associated with high error rates. Distributed cognition in medical record abstraction has not been studied as a possible explanation for abstraction errors. We employed the theory of distributed representation and representational analysis to systematically evaluate cognitive demands in medical record abstraction and the extent of external cognitive support employed in a sample of clinical research data collection forms. We show that the cognitive load required for abstraction in 61% of the sampled data elements was high, exceedingly so in 9%. Further, the data collection forms did not support external cognition for the most complex data elements. High working memory demands are a possible explanation for the association of data errors with data elements requiring abstractor interpretation, comparison, mapping or calculation. The representational analysis used here can be used to identify data elements with high cognitive demands.

An evaluation of the centers for Medicare & Medicaid Services' Hospital Acquired Conditions and Present on Admission Indicator Reporting program

Objective: This dissertation evaluated three aspects of the Centers for Medicare and Medicaid Services' Hospital Acquired Conditions and Present on Admission Indicator Reporting program (HACPOA program) to produce three journal articles for publication. ^ Methods: All payer admission records from state inpatient databases from Arizona, New Jersey and Washington states were analyzed for the year 2008. However some analyses required a sample of adult only Medicare patients in the first two studies. California's inpatient data (2004 – 2010) was also analyzed in the third study to examine the reporting and non-payment program elements' impact on the incidence of hospital acquired conditions. ^ Results: Majority diagnoses reported in inpatient prospective payment systems hospitals were present on admission. However, some diagnoses are still coded as "not present on admission" and "insufficient documentation to determine whether or not conditions are present on admission or not". This is important because it reveals that hospital complications still occur in hospitals. Hospital fall and trauma injuries were the most common hospital acquired conditions observed in this study. Predictors of hospital fall injuries include age, gender, number of diagnoses, number of procedures, number of chronic conditions while predictors of hospital trauma injuries include number of e-codes, number of diagnoses and the presence of chronic conditions on a patient's admission records. Finally, the implementation of the present on admission reporting requirement increased reports of certain hospital acquired conditions while the non-payment policy element in the Hospital Acquired Conditions program reduced the incidence of hospital fall and trauma injuries in particular. ^ Conclusion: The implementation of the Hospital Acquired Conditions and Present on Admission Indicator Reporting program has made the state inpatient database a more useful source of data capable of now identifying hospital complications. The reporting and nonpayment program elements in the HACPOA program have also impacted the incidence of hospital acquired conditions. ^

Data quality in trauma transfusion studies and the impact of missing data on predicting massive transfusion

Maximizing data quality may be especially difficult in trauma-related clinical research. Strategies are needed to improve data quality and assess the impact of data quality on clinical predictive models. This study had two objectives. The first was to compare missing data between two multi-center trauma transfusion studies: a retrospective study (RS) using medical chart data with minimal data quality review and the PRospective Observational Multi-center Major Trauma Transfusion (PROMMTT) study with standardized quality assurance. The second objective was to assess the impact of missing data on clinical prediction algorithms by evaluating blood transfusion prediction models using PROMMTT data. RS (2005-06) and PROMMTT (2009-10) investigated trauma patients receiving ≥ 1 unit of red blood cells (RBC) from ten Level I trauma centers. Missing data were compared for 33 variables collected in both studies using mixed effects logistic regression (including random intercepts for study site). Massive transfusion (MT) patients received ≥ 10 RBC units within 24h of admission. Correct classification percentages for three MT prediction models were evaluated using complete case analysis and multiple imputation based on the multivariate normal distribution. A sensitivity analysis for missing data was conducted to estimate the upper and lower bounds of correct classification using assumptions about missing data under best and worst case scenarios. Most variables (17/33=52%) had <1% missing data in RS and PROMMTT. Of the remaining variables, 50% demonstrated less missingness in PROMMTT, 25% had less missingness in RS, and 25% were similar between studies. Missing percentages for MT prediction variables in PROMMTT ranged from 2.2% (heart rate) to 45% (respiratory rate). For variables missing >1%, study site was associated with missingness (all p≤0.021). Survival time predicted missingness for 50% of RS and 60% of PROMMTT variables. MT models complete case proportions ranged from 41% to 88%. Complete case analysis and multiple imputation demonstrated similar correct classification results. Sensitivity analysis upper-lower bound ranges for the three MT models were 59-63%, 36-46%, and 46-58%. Prospective collection of ten-fold more variables with data quality assurance reduced overall missing data. Study site and patient survival were associated with missingness, suggesting that data were not missing completely at random, and complete case analysis may lead to biased results. Evaluating clinical prediction model accuracy may be misleading in the presence of missing data, especially with many predictor variables. The proposed sensitivity analysis estimating correct classification under upper (best case scenario)/lower (worst case scenario) bounds may be more informative than multiple imputation, which provided results similar to complete case analysis.^