The economic burden of genetic tests for the infertile male: a pilot algorithm to improve test predictive value.

PURPOSE: We developed a model to optimize genetic testing in infertile men with nonobstructive azoospermia and severe oligospermia. We also assessed the optimal cutoff value of the predicted probability of advising genetic testing and evaluated the direct cost saving of using the model. MATERIALS AND METHODS: We retrospectively reviewed the records of infertile men who underwent Y microdeletion and karyotype testing at our fertility center from 2006 to 2012. Semen parameters, testicular volume, testosterone, luteinizing hormone, follicular stimulating hormone and varicocele were assessed as potential predictors of genetic disorders. We fitted logistic regression to all predictors and selected a nomogram based on the concordance index and calibration. We calculated the cost saving of using the model. RESULTS: Of 325 patients 278 fulfilled study inclusion criteria, including 27 with an abnormal karyotype, 11 with a Y microdeletion and 1 with each condition. We developed a nomogram using sperm concentration and motility, testicular volume and serum testosterone level. The nomogram concordance index was 0.738. The optimal cutoff value was 13.8% with 0.788 sensitivity, 0.590 specificity, 0.245 positive predictive value and 0.943 negative predictive value. Testing men above the 13.8% cutoff resulted in a direct 45% cost saving. However, 15.4% of genetic anomalies were missed, including 2 Y microdeletions. CONCLUSIONS: Using common clinical and laboratory parameters our nomogram detects 84.6% of genetic anomalies. Nomogram use resulted in a 45% direct cost saving but carries the risk of missing pertinent genetic abnormalities.

Duke Scholars

Author Karen Christine Baker Urology