Validity of an unsupervised self-administered questionnaire for self-assessment of melanoma risk

The Level of Agreement between Self-Assessments and Examiner Assessments of Melanocytic Nevus Counts: Findings from an Evaluation of 4548 Double Assessments

Cutaneous melanoma (CM) is a candidate for screening programs because its prognosis is excellent when diagnosed at an early disease stage. Targeted screening of those at high risk for developing CM, a cost-effective alternative to population-wide screening, requires valid procedures to identify the high-risk group. Self-assessment of the number of nevi has been suggested as a component of such procedures, but its validity has not yet been established. We analyzed the level of agreement between self-assessments and examiner assessments of the number of melanocytic nevi in the area between the wrist and the shoulder of both arms based on 4548 study subjects in whom mutually blinded double counting of nevi was performed. Nevus counting followed the IARC protocol. Study subjects received written instructions, photographs, a mirror, and a "nevometer" to support self-assessment of nevi larger than 2 mm. Nevus counts were categorized based on the quintiles of the distribution into five levels, defining a nevus score. Cohen's weighted kappa coefficient (κ) was estimated to measure the level of agreement. In the total sample, the agreement between self-assessments and examiner assessments was moderate (weighted κ = 0.596). Self-assessed nevus counts were higher than those determined by trained examiners (mean difference: 3.33 nevi). The level of agreement was independent of sociodemographic and cutaneous factors; however, participants' eye color had a significant impact on the level of agreement. Our findings show that even with comprehensive guidance, only a moderate level of agreement between self-assessed and examiner-assessed nevus counts can be achieved. Self-assessed nevus information does not appear to be reliable enough to be used in individual risk assessment to target screening activities.

Inter-Rater Agreement in Assessing Risk of Bias in Melanoma Prediction Studies Using the Prediction Model Risk of Bias Assessment Tool (PROBAST): Results from a Controlled Experiment on the Effect of Specific Rater Training

Assessing the risk of bias (ROB) of studies is an important part of the conduct of systematic reviews and meta-analyses in clinical medicine. Among the many existing ROB tools, the Prediction Model Risk of Bias Assessment Tool (PROBAST) is a rather new instrument specifically designed to assess the ROB of prediction studies. In our study we analyzed the inter-rater reliability (IRR) of PROBAST and the effect of specialized training on the IRR. Six raters independently assessed the risk of bias (ROB) of all melanoma risk prediction studies published until 2021 (n = 42) using the PROBAST instrument. The raters evaluated the ROB of the first 20 studies without any guidance other than the published PROBAST literature. The remaining 22 studies were assessed after receiving customized training and guidance. Gwet's AC₁ was used as the primary measure to quantify the pairwise and multi-rater IRR. Depending on the PROBAST domain, results before training showed a slight to moderate IRR (multi-rater AC₁ ranging from 0.071 to 0.535). After training, the multi-rater AC₁ ranged from 0.294 to 0.780 with a significant improvement for the overall ROB rating and two of the four domains. The largest net gain was achieved in the overall ROB rating (difference in multi-rater AC₁: 0.405, 95%-CI 0.149-0.630). In conclusion, without targeted guidance, the IRR of PROBAST is low, questioning its use as an appropriate ROB instrument for prediction studies. Intensive training and guidance manuals with context-specific decision rules are needed to correctly apply and interpret the PROBAST instrument and to ensure consistency of ROB ratings.

Factors associated with suspected nonmelanoma skin cancers, dysplastic nevus, and cutaneous melanoma among first-time SpotMe screening program participants during 2009-2010

BACKGROUND

There have been no studies of the American Academy of Dermatology's SpotMe skin cancer screening program to collectively analyze and determine the factors associated with suspected basal cell carcinoma (BCC), squamous cell carcinoma (SCC), dysplastic nevus (DN), and cutaneous melanoma (CM) diagnoses.

OBJECTIVE

Describe the demographics, risk factors, and access to care profiles associated with suspected diagnoses of BCC, SCC, DN, and CM among first-time SpotMe screenees during 2009-2010.

METHODS

We conducted a cross-sectional analysis of data from the SpotMe skin cancer screenings conducted in 2009 and 2010. We performed multivariable logistic regression analysis for each diagnosis, incorporating standard demographic, access to care, and risk factor variables in the models.

RESULTS

Men, those without a regular dermatologist, persons reporting recently changing moles, and those with a personal history of melanoma were at increased risk for each of the suspected diagnoses analyzed. Uninsured persons were at increased risk for suspected malignancies (BCC, SCC, and CM).

LIMITATIONS

Lack of histologic confirmation for diagnoses and cross-sectional design.

CONCLUSION

Among first-time SpotMe participants, suspected diagnoses of BCC, SCC, DN, and CM shared several associated factors, which may be considered when planning outreach and screening for populations at risk for skin cancer.

Using the Prediction Model Risk of Bias Assessment Tool (PROBAST) to Evaluate Melanoma Prediction Studies

Simple Summary

The rising incidence of cutaneous melanoma over recent decades, combined with a general interest in cancer risk prediction, has led to a high number of published melanoma risk prediction models. The aim of our work was to assess the validity of these models in order to discuss the current state of knowledge about how to predict incident cutaneous melanoma. To assess the risk of bias, we used a standardized procedure based on PROBAST (Prediction model Risk Of Bias ASsessment Tool). Only one of the 42 studies identified was rated as having a low risk of bias. However, it was encouraging to observe a recent reduction of problematic statistical methods used in the analyses. Nevertheless, the evidence base of high-quality studies that can be used to draw conclusions on the prediction of incident cutaneous melanoma is currently much weaker than the high number of studies on this topic would suggest.

Abstract

Rising incidences of cutaneous melanoma have fueled the development of statistical models that predict individual melanoma risk. Our aim was to assess the validity of published prediction models for incident cutaneous melanoma using a standardized procedure based on PROBAST (Prediction model Risk Of Bias ASsessment Tool). We included studies that were identified by a recent systematic review and updated the literature search to ensure that our PROBAST rating included all relevant studies. Six reviewers assessed the risk of bias (ROB) for each study using the published “PROBAST Assessment Form” that consists of four domains and an overall ROB rating. We further examined a temporal effect regarding changes in overall and domain-specific ROB rating distributions. Altogether, 42 studies were assessed, of which the vast majority (n = 34; 81%) was rated as having high ROB. Only one study was judged as having low ROB. The main reasons for high ROB ratings were the use of hospital controls in case-control studies and the omission of any validation of prediction models. However, our temporal analysis results showed a significant reduction in the number of studies with high ROB for the domain “analysis”. Nevertheless, the evidence base of high-quality studies that can be used to draw conclusions on the prediction of incident cutaneous melanoma is currently much weaker than the high number of studies on this topic would suggest.

Development and validation of a self-management behavior questionnaire for Chinese enterostomy patients

Objectives

This study aimed to develop a self-management behavior questionnaire for Chinese enterostomy patients and examine its reliability and validity.

Methods

Guided by the theory of self-management, an initial questionnaire was generated through literature review, group meetings, and two rounds of an expert consultation. Finally, the reliability and validity of the questionnaire were validated through a questionnaire survey of 200 enterostomy patients were selected from the Affiliated Hospital of Medical University from June 2016 to March 2017.

Results

The content validity index was 0.80–1.00. The exploratory factor analysis yielded a five-factor (dietary behavior, psychosocial behavior, symptom management behavior, medical compliance behavior, information management behavior), consisting of 40 items. The cumulative variance contribution rate was 65.42%. The Cronbach’s α coefficient for the total questionnaire was 0.972, and for the five factors ranged from 0.797 to 0.939, indicating a good internal consistency. The test-retest reliability was 0.867 (P < 0.01). The self-management behavior questionnaire score was negatively related to the Ostomy Skin Tool score (r = −0.800, P < 0.01).

Conclusions

The self-management behavior questionnaire developed in this study showed good reliability and validity and can be used to assess the self-management behavior of Chinese enterostomy patients.

Reporting Quality of Studies Developing and Validating Melanoma Prediction Models: An Assessment Based on the TRIPOD Statement

Transparent and accurate reporting is essential to evaluate the validity and applicability of risk prediction models. Our aim was to evaluate the reporting quality of studies developing and validating risk prediction models for melanoma according to the TRIPOD (Transparent Reporting of a multivariate prediction model for Individual Prognosis Or Diagnosis) checklist. We included studies that were identified by a recent systematic review and updated the literature search to ensure that our TRIPOD rating included all relevant studies. Six reviewers assessed compliance with all 37 TRIPOD components for each study using the published “TRIPOD Adherence Assessment Form”. We further examined a potential temporal effect of the reporting quality. Altogether 42 studies were assessed including 35 studies reporting the development of a prediction model and seven studies reporting both development and validation. The median adherence to TRIPOD was 57% (range 29% to 78%). Study components that were least likely to be fully reported were related to model specification, title and abstract. Although the reporting quality has slightly increased over the past 35 years, there is still much room for improvement. Adherence to reporting guidelines such as TRIPOD in the publication of study results must be adopted as a matter of course to achieve a sufficient level of reporting quality necessary to foster the use of the prediction models in applications.

Eruptive Cherry Angiomas and Skin Melanoma: Further Insights into an Intriguing Association

BACKGROUND

Some risk factors for malignant melanoma (MM) are recognized.

OBJECTIVE

To compare the strength of association between MM and eruptive cherry angiomas (CAs) with that of other well-known associations.

METHODS

This cross-sectional study included all subjects referred to the Outpatient Dermatology-Oncology and Dermoscopy Units of the Universities of Ferrara and Bologna, Italy, over a 5-month period and submitted to total body skin examination. We recorded: age, sex, cutaneous and non-cutaneous malignancies, presence of CAs, arbitrarily considered as "eruptive" when >10, >40 common melanocytic naevi or >2 clinically atypical naevi. The strength of association between the possible risk factors and MM was calculated by odds ratio in both the whole population and age quartiles. Variables associated with MM were included in multiple logistic regression analysis.

RESULTS

1,190 subjects were included; 615 had malignant skin tumours, 462 MM, 85 extracutaneous tumours. Five hundred and eighty-seven subjects had eruptive CAs, 485 subjects >40 melanocytic naevi and 368 more than 2 atypical melanocytic naevi. Eruptive CAs, especially in subjects younger than 70, and >2 atypical melanocytic naevi, mostly in subjects older than 50, were significantly associated with MM. The strength of these 2 associations was similar. The presence of >40 melanocytic naevi was not associated with MM.

CONCLUSIONS

These findings confirmed an association between MM and eruptive CAs, which was as strong as the one between MM and >2 atypical melanocytic naevi. CAs seem an intriguing model of interaction between heterogeneous variables, like immunocompetence, stimuli inducing endothelial cell proliferation, and oncogenesis, which deserves further investigation.

Risk Prediction Models for Melanoma: A Systematic Review on the Heterogeneity in Model Development and Validation

The rising incidence of cutaneous melanoma over the past few decades has prompted substantial efforts to develop risk prediction models identifying people at high risk of developing melanoma to facilitate targeted screening programs. We review these models, regarding study characteristics, differences in risk factor selection and assessment, evaluation, and validation methods. Our systematic literature search revealed 40 studies comprising 46 different risk prediction models eligible for the review. Altogether, 35 different risk factors were part of the models with nevi being the most common one (n = 35, 78%); little consistency in other risk factors was observed. Results of an internal validation were reported for less than half of the studies (n = 18, 45%), and only 6 performed external validation. In terms of model performance, 29 studies assessed the discriminative ability of their models; other performance measures, e.g., regarding calibration or clinical usefulness, were rarely reported. Due to the substantial heterogeneity in risk factor selection and assessment as well as methodologic aspects of model development, direct comparisons between models are hardly possible. Uniform methodologic standards for the development and validation of risk prediction models for melanoma and reporting standards for the accompanying publications are necessary and need to be obligatory for that reason.

Strategizing Screening for Melanoma in an Era of Novel Treatments: A Model-Based Approach

BACKGROUND

Benefit-harm tradeoffs of melanoma screening depend on disease risk and treatment efficacy. We developed a model to project outcomes of screening for melanoma in populations with different risks under historic and novel systemic treatments.

METHODS

Computer simulation model of a screening program with specified impact on overall and advanced-stage incidence. Inputs included meta-analyses of treatment trials, cancer registry data, and a melanoma risk prediction study RESULTS: Assuming 50% reduction in advanced stage under screening, the model projected 59 and 38 lives saved per 100,000 men under historic and novel treatments, respectively. With 10% increase in stage I, the model projects 2.9 and 4.7 overdiagnosed cases per life saved and number needed to be screened (NNS) equal to 1695 and 2632 under historical and novel treatments. When screening was performed only for the 20% of individuals with highest predicted risk, 34 and 22 lives per 100,000 were saved under historic and novel treatments. Similar results were obtained for women, but lives saved were lower.

CONCLUSIONS

Melanoma early detection programs must shift a substantial fraction of cases from advanced to localized stage to be sustainable. Advances in systemic therapies for melanoma might noticeably reduce benefits of screening, but restricting screening to individuals at highest risk will likely reduce intervention efforts and harms while preserving >50% of the benefit of nontargeted screening.

IMPACT

Our accessible modeling framework will help to guide population melanoma screening programs in an era of novel treatments for advanced disease.

Relative Importance of Traditional Risk Factors for Malignant Melanoma in the Czech Population

AIM

The Czech Republic is one of the leading European countries in incident cases of malignant melanoma (MM), which is on the rise. The study objective was to assess the strength of associations between MM and the known generally accepted risk factors for MM in the population of the Czech Republic.

METHODS

The study was designed as a case-control study where cases were incident cases of MM detected at the Department of Dermatology and Venereology of the Bulovka Hospital. Controls were selected from cancer-free patients admitted to departments other than Dermatology and Venereology. Validated questionnaires were used to collect demographic, epidemiological, and clinical data.

RESULTS

The binary logistic model shows the main risk factors for MM: male, female (OR=0.292, 95% CI=0.175-0.486), a changed mole (OR=6.371, 95% CI=3.774-10.756), a history of skin cancer (OR=95.704, 95% CI=37.241-10.756), and sunbeds use (OR=3.594, 95% CI=1.288-10.028). Using sunscreen products was considered as a protective factor against MM (OR=0.253, 95% CI=0.137-0.466).

CONCLUSION

The primary and secondary prevention increasingly emerges as a public health priority in the effort to reverse the negative trend in cases of MM and mortality from this disease in the Czech Republic. A prerequisite for an effective secondary prevention through screening is, among others, the identification of the population groups at highest risk for MM.

Independent validation of six melanoma risk prediction models

Identifying people at high risk of melanoma is important for targeted prevention activities and surveillance. Several tools have been developed to classify melanoma risk, but few have been independently validated. We assessed the discriminatory performance of six melanoma prediction tools by applying them to individuals from two independent data sets, one comprising 762 melanoma cases and the second a population-based sample of 42,116 people without melanoma. We compared the model predictions with actual melanoma status to measure sensitivity and specificity. The performance of the models was variable with sensitivity ranging from 97.7 to 10.5% and specificity from 99.6 to 1.3%. The ability of all the models to discriminate between cases and controls, however, was generally high. The model developed by MacKie et al. (1989) had higher sensitivity and specificity for men (0.89 and 0.88) than women (0.79 and 0.72). The tool developed by Cho et al. (2005) was highly specific (men, 0.92; women, 0.99) but considerably less sensitive (men, 0.64; women, 0.37). Other models were either highly specific but lacked sensitivity or had low to very low specificity and higher sensitivity. Poor performance was partly attributable to the use of non-standardized assessment items and various differing interpretations of what constitutes "high risk".

Predicting melanoma risk: theory, practice and future challenges

The incidence of melanoma continues to rise in most fair-skinned populations. Strategies to curb the toll from melanoma include targeting the patients who are at highest risk with the aim of either preventing the onset of cancer or intervening early in order to improve survival. The challenge has been to synthesize the available information on risk factors into prediction tools with clinical utility, such that 'high-risk' patients can be identified with accuracy. While a number of risk prediction tools for melanoma have been developed, few have undergone rigorous evaluation of their performance in order to assess calibration or discrimination, and even fewer have been validated in independent populations. Future research should assess the validity of existing tools and seek to integrate the increasing volumes of data being generated by genomic studies.

Risk prediction models for melanoma: a systematic review

Melanoma incidence is increasing rapidly worldwide among white-skinned populations. Earlier diagnosis is the principal factor that can improve prognosis. Defining high-risk populations using risk prediction models may help targeted screening and early detection approaches. In this systematic review, we searched Medline, EMBASE, and the Cochrane Library for primary research studies reporting or validating models to predict risk of developing cutaneous melanoma. A total of 4,141 articles were identified from the literature search and six through citation searching. Twenty-five risk models were included. Between them, the models considered 144 possible risk factors, including 18 measures of number of nevi and 26 of sun/UV exposure. Those most frequently included in final risk models were number of nevi, presence of freckles, history of sunburn, hair color, and skin color. Despite the different factors included and different cutoff values for sensitivity and specificity, almost all models yielded sensitivities and specificities that fit along a summary ROC with area under the ROC (AUROC) of 0.755, suggesting that most models had similar discrimination. Only two models have been validated in separate populations and both also showed good discrimination with AUROC values of 0.79 (0.70-0.86) and 0.70 (0.64-0.77). Further research should focus on validating existing models rather than developing new ones.

Individual risk of cutaneous melanoma in New Zealand: developing a clinical prediction aid

Background

New Zealand and Australia have the highest melanoma incidence rates worldwide. In New Zealand, both the incidence and thickness have been increasing. Clinical decisions require accurate risk prediction but a simple list of genetic, phenotypic and behavioural risk factors is inadequate to estimate individual risk as the risk factors for melanoma have complex interactions. In order to offer tailored clinical management strategies, we developed a New Zealand prediction model to estimate individual 5-year absolute risk of melanoma.

Methods

A population-based case–control study (368 cases and 270 controls) of melanoma risk factors provided estimates of relative risks for fair-skinned New Zealanders aged 20–79 years. Model selection techniques and multivariate logistic regression were used to determine the important predictors. The relative risks for predictors were combined with baseline melanoma incidence rates and non-melanoma mortality rates to calculate individual probabilities of developing melanoma within 5 years.

Results

For women, the best model included skin colour, number of moles > =5 mm on the right arm, having a 1st degree relative with large moles, and a personal history of non-melanoma skin cancer (NMSC). The model correctly classified 68% of participants; the C-statistic was 0.74. For men, the best model included age, place of occupation up to age 18 years, number of moles > =5 mm on the right arm, birthplace, and a history of NMSC. The model correctly classified 67% of cases; the C-statistic was 0.71.

Conclusions

We have developed the first New Zealand risk prediction model that calculates individual absolute 5-year risk of melanoma. This model will aid physicians to identify individuals at high risk, allowing them to individually target surveillance and other management strategies, and thereby reduce the high melanoma burden in New Zealand.

MC1R genotype as a predictor of early-onset melanoma, compared with self-reported and physician-measured traditional risk factors: an Australian case-control-family study

Background

Melanocortin-1 receptor (MC1R) gene variants are very common and are associated with melanoma risk, but their contribution to melanoma risk prediction compared with traditional risk factors is unknown. We aimed to 1) evaluate the separate and incremental contribution of MC1R genotype to prediction of early-onset melanoma, and compare this with the contributions of physician-measured and self-reported traditional risk factors, and 2) develop risk prediction models that include MC1R, and externally validate these models using an independent dataset from a genetically similar melanoma population.

Methods

Using data from an Australian population-based, case-control-family study, we included 413 case and 263 control participants with sequenced MC1R genotype, clinical skin examination and detailed questionnaire. We used unconditional logistic regression to estimate predicted probabilities of melanoma. Results were externally validated using data from a similar study in England.

Results

When added to a base multivariate model containing only demographic factors, MC1R genotype improved the area under the receiver operating characteristic curve (AUC) by 6% (from 0.67 to 0.73; P < 0.001) and improved the quartile classification by a net 26% of participants. In a more extensive multivariate model, the factors that contributed significantly to the AUC were MC1R genotype, number of nevi and previous non-melanoma skin cancer; the AUC was 0.78 (95% CI 0.75-0.82) for the model with self-reported nevi and 0.83 (95% CI 0.80-0.86) for the model with physician-counted nevi. Factors that did not further contribute were sun and sunbed exposure and pigmentation characteristics. Adding MC1R to a model containing pigmentation characteristics and other self-reported risk factors increased the AUC by 2.1% (P = 0.01) and improved the quartile classification by a net 10% (95% CI 1-18%, P = 0.03).

Conclusions

Although MC1R genotype is strongly associated with skin and hair phenotype, it was a better predictor of early-onset melanoma than was pigmentation characteristics. Physician-measured nevi and previous non-melanoma skin cancer were also strong predictors. There might be modest benefit to measuring MC1R genotype for risk prediction even if information about traditional self-reported or clinically measured pigmentation characteristics and nevi is already available.

Identifying Persons at Highest Risk of Melanoma Using Self-Assessed Risk Factors

OBJECTIVE: To develop a self-assessed melanoma risk score to identify high-risk persons for screening METHODS: We used data from a 1997 melanoma case-control study from Washington State, USA, where 386 cases with invasive cutaneous melanoma and 727 controls were interviewed by telephone. A logistic regression prediction model was developed on 75% of the data and validated in the remaining 25% by calculating the area under the receiver operating characteristic curve (AUC), a measure of predictive accuracy from 0.5-1 (higher scores indicating better prediction). A risk score was calculated for each individual, and sensitivities for various risk cutoffs were calculated. RESULTS: The final model included sex, age, hair color, density of freckles, number of severe sunburns in childhood and adolescence, number of raised moles on the arms, and history of non-melanoma skin cancer. The area under the receiver operating characteristic curve(AUC) was 0.70 (95% CI: 0.64, 0.77). The top 15% risk group included 50% of melanomas (sensitivity 50%). CONCLUSIONS: This self-assessed score could be used as part of a comprehensive melanoma screening and public education program to identify high-risk individuals in the general population. This study suggests it may be possible to capture a large proportion of melanomas by screening a small high-risk group. Further study is needed to determine the costs, feasibility, and risks of this approach.

Defining the patient at high risk for melanoma

In this practical review, we aim to help clinicians identify patients who are at significant risk of developing malignant melanoma. Universal screening is challenging, thus it is important to effectively single out patients who have a high risk of developing the disease. We provide a summary of pertinent questions to review when taking the patient's history, point out the phenotypic features to note during skin examination, and suggest risk stratification as a means to plan initial and long-term surveillance strategy. We mention personal and family history of melanoma as prime risk factors for melanoma, yet the review also focuses on the patient who has no history of melanoma, either in himself or his family, and the proper ways to evaluate his likelihood of developing the disease.

Creation and test of a questionnaire for self-assessment of melanoma risk factors

The objective of this study was to create a self-administrated questionnaire for people to enable them to assess their own melanoma risk factors. To test the validity of this questionnaire in a large prospective study, the answers given by the patient were systematically checked by his or her general practitioner. In this prospective study, the choice of questions was based on a review of the literature. The validity of the questionnaire was assessed by testing 1500 consecutive patients attending a consultation with their general practitioner. Considerable variations concerning the prevalence of different melanoma risk factors were noticed in the population: 44.1% had a phototype I or II, 41% had severe sunburn during infancy, 29.9% had freckling tendency, 22% had more than 50 naevi and 1.4% a personal history of melanoma. In total, 45% had more than one melanoma risk factor. The accuracy of the answers given by the patients was assured by the correction given by their general practitioners. The percentage of correct answers given by the patients was 79.9% for the phototype, 90.6% for freckling tendency, 86.6% for the number of naevi, 96.5% for severe sunburn during infancy and 98.1 and 95.8% for personal and familial history of melanoma. This study confirms that individuals with multiple risk factors for melanoma are common among patients consulting their general practitioners. Furthermore, self-screening with the self-assessment questionnaire is easily feasible and is accurate for identifying high-risk individuals. This tool might be useful for carrying out melanoma-targeted screening.

Evaluating the usefulness of self-reported risk factors in a skin cancer screening program

The objective of this study was to describe the risk factor profile of skin cancer screening participants and to determine whether there is an association between the number of skin cancer/melanoma risk factors and the likelihood of diagnosis of a malignant melanoma. Seventy skin cancer screening clinics were held by the Lions Cancer Institute in predominantly rural areas of Western Australia between 1996 and 2003. Participants were self-selected and voluntary, responding to an advertisement seeking people at 'high-risk' of melanoma. The Lions Cancer Institute skin screening clinics targeted participation by individuals with three or more of the established risk factors for skin cancer/melanoma. Questionnaires collecting information on the self-report of nine risk factors were completed by 5950 participants who were screened for melanoma between 1996 and 2003. The number and type of risk factors, and of provisionally diagnosed and histopathologically confirmed malignant melanomas were measured. Of 5950 participants, 18 histopathologically confirmed malignant melanomas were detected. A participant's total number of risk factors showed some association with the provisional melanoma diagnosis given at the time of screening. No relationship, however, was observed between the number of risk factors and a melanoma that was histopathologically confirmed after screening. The risk factor method is effective in selecting a 'high risk' population, but does not seem to have high value in predicting who will be diagnosed with melanoma as a result of screening. Further studies are needed to verify this finding owing to the rarity of melanoma and the small number of confirmed melanomas in this study.

Nomographic representation of logistic regression models: a case study using patient self-assessment data

Logistic regression models are widely used in medicine, but difficult to apply without the aid of electronic devices. In this paper, we present a novel approach to represent logistic regression models as nomograms that can be evaluated by simple line drawings. As a case study, we show how data obtained from a questionnaire-based patient self-assessment study on the risks of developing melanoma can be used to first identify a subset of significant covariates, build a logistic regression model, and finally transform the model to a graphical format. The advantage of the nomogram is that it can easily be mass-produced, distributed and evaluated, while providing the same information as the logistic regression model it represents.