Institutional management of patients with erectile dysfunction
This study was performed in a secondary referral center for endocrine diseases and metabolic disorders accredited by the Italian Society of Andrology and Sexual Medicine (SIAMS), the Outpatients Clinic of Endocrinology, and Metabolic Disease of Conversano Hospital (Italy). Accordingly, patients with sexual and reproductive disorders are referred for the initial and long-term management of male hypogonadism, ED, ejaculation dysfunction, and infertility.
In individuals complaining of ED, the initial evaluation includes the assessment of 1) detailed medical and sexual history, including the evaluation of smoking status; 2) physical examination and anthropometry, including blood pressure, body weight, height, body mass index (BMI), and waist circumference (WC); 3) laboratory testing, including fasting plasma glucose, HbA1c, lipid profile, androgens and gonadotropins (LH, FSH); 4) questionnaires (i.e., IIEF-5). Total testosterone (TT) is assessed in all; in patients with medical conditions affecting the sex hormone-binding globulin (SHBG) levels or whose initial TT concentrations are at or near the lower limit of the normal range (e.g., 300 ng/dl [10.4 nM/L]), SHBG [14]. The IIEF-5, also known as the Sexual Health Inventory for Men (SHIM), is used to classify the severity of erectile dysfunction [15]. Additional procedures were performed when indicated. Given the high prevalence of cardiovascular disease, a multidisciplinary diagnostic workup is considered. Other laboratory assessments (e.g., PSA, prolactin) and examination (e.g. workup for obstructive sleep apnea) are assessed in subjects with specific signs and associated symptoms. Different procedures (multiple pituitary hormone assessments, genetic analysis, magnetic resonance imaging) are performed in subjects with TT levels < 150 ng/dl (5.2 nM/L), given that the risk of an organic form of hypogonadism in this subgroup is high. This approach is based on our previous experience of over more than three decades and is consistent with the most recent literature [16–18].
Study design
This was a retrospective study. The study period was from January 1, 2018, to December 31, 2019. All consecutive adult male patients complaining of ED, with available data on TT, IIEF-5 questionnaire, and comorbidities, were initially eligible for the study and were searched in the institutional database. Patients were excluded only if the abovementioned data could not be retrieved. All procedures were according to the ethical standards of the University of Bari (protocol number 6454, July 2020) and the Azienda Sanitaria Locale (protocol number: 1294, October 2020) Declaration of Helsinki. To be included in the study, all participants signed the informed consent.
Procedures
Data collected during the initial evaluation were used. Each patient at the Outpatient Clinic completed the IIEF-5 questionnaire. ED was defined as IIEF-5 score ≤ 21 and was classified as severe for scores 5–7, moderate for scores 8–11, mild to moderate for scores 12–16, and mild for 17–21 [16]. Self-reported or newly diagnosed comorbidities were used to calculate the CCI score. Clinical conditions and associated scores were as follows. Cerebrovascular disease, chronic lung disease, congestive heart failure, connective tissue disease, dementia, diabetes, mild liver disease, myocardial infarction, peripheral vascular disease, and ulcer were assigned 1 point each. Any tumor, diabetes with end-organ damage, hemiplegia, leukemia, lymphoma, and moderate or severe kidney disease (eGFR < 60 ml/min) were assigned 2 points. The moderate or severe liver diseases were assigned 3 points each. Acquired immunodeficiency syndrome (AIDS) and solid metastatic tumor were assigned 6 points each [13]. Despite being acknowledged risk factors for ED, arterial hypertension, dyslipidemia, and tobacco smoking had been not initially included in the CCI. Therefore, a modified index (modified Charlson comorbidity index [mCCI]) was developed; the abovementioned disorders were included and assigned 1-point each [4, 18].
Laboratory measurements
Hormonal parameters were exclusively determined in the laboratory of our hospital as for habitual clinical practice. Each blood specimen was collected between 8, and 9 am after an overnight fast and stored at -20°C until analyzed. Serum LH, FSH, TT, and SHBG were measured by commercial immunometric assays (Immulite, EURO/DPC, UK). Reference range were 7.5 ± 2.6 IU/L for LH, 6.6 ± 2.5 IU/L for FSH, 450 ± 90 ng/dL for TT (15.6 ± 3.1 nM/L) and 45.4 ± 5.1 nM/L for SHBG. Intra- and inter-assay coefficients of variation of these methods were < 8% and < 10%, respectively [20].
Statistical analysis
All analyses were conducted using SAS 9.4 software (SAS Institute, Cary, NC, USA).
The primary outcome was to assess the prevalence of ED according to the level of severity of CCI. The secondary outcomes included the correlation between 1) IIEF-5 and TT; 2) CCI and TT; 3) IIEF-5 and CCI. Finally, the performance of the CCI and mCCI were compared. Continuous variables were expressed as mean and standard deviation (SD) for normally distributed parameters or the median and interquartile range (IQR) in skewed data distribution. Shapiro-Wilk’s statistics were used to test normality, and an appropriate function was applied to transform those showing a non-normal distribution. The distribution of patients in each category was described as frequency and proportion. Chi-square test and Chi-squared test for trend were used to verify the association between ED and CCI grouped in four classes (score = 0, score = 1, score = 2, score ≥ 3). Pearson's correlation coefficient or Spearman’s non-parametric correlation coefficient was used to testing the secondary outcomes, as necessary, and the Mardia test was used to verify multivariate normality. Univariate and multivariable linear models were applied to evaluate the effect of the parameters age, BMI, TT, LH, FSH, SHBG, albumin, and CCI on the IIEF-5 scores. The R-squared value and the residual normality test were used to evaluate the goodness of fit of the multivariable model. All tests of statistical significance were two-tailed, and p-values less than 0.05 were considered statistically significant.