PURPOSE
To evaluate the association between phosphodiesterase type-5 inhibitor (PDE-5i) therapy and the development of glaucoma suspect (GS) status or open-angle glaucoma (OAG) in men with erectile dysfunction (ED).
DESIGN
Multicenter, retrospective cohort study using deidentified electronic health records.
SUBJECTS
Male adults aged ≥ 40 years with ED, with or without chronic PDE-5i exposure.
METHODS
We used propensity score matching (PSM) to balance demographics, systemic and ophthalmic comorbidities, medications, laboratory measures. Additionally, we applied Cox proportional models on the unmatched cohorts to estimate adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs).
MAIN OUTCOME MEASURES
Development of GS or OAG at 1, 2, and 3-years following PDE-5i exposure.
RESULTS
Before matching, 40,676 men received PDE-5i therapy and 32,415 had no exposure. After PSM, the development of GS status was less frequent among PDE-5i users at 1 year (6.49% vs 9.73%, P <.01), 2 years (9.27% vs 10.86%, P <.01), and 3 years (11.17% vs 12.06%, P =.01). Similarly, the development of OAG was less frequent among PDE-5i users at 1 year (2.13% vs 3.22%, P <.01), 2 years (3.00% vs 3.80%, P <.01), and 3 years (3.88% vs 4.28%, P =.04). Cox proportional models demonstrated reduced hazards for GS at 1 year (aHR 0.78; 95% C: 0.75-0.81; P <.01), 2 years (0.89; 0.86-0.93; P <.01), and 3 years (0.94; 0.90-0.97; P =.01), and for OAG at 1 year (0.76; 0.71-0.81; P <.01), 2 years (0.79; 0.75-0.83; P <.01), and 3 years (0.84; 0.79-0.88; P =.01).
CONCLUSION
Chronic PDE-5i use was associated with a lower hazard of developing GS and OAG over 1-, 2-, and 3-years of follow-up. These findings suggest a potential protective association between PDE-5i and glaucoma development that requires confirmation in prospective studies.
INTRODUCTION
G laucoma is a chronic, progressive optic neuropathy characterized by degeneration of retinal ganglion cells (RGCs) and their axons, leading to irreversible vision loss. Although its exact pathogenesis remains unclear, glaucoma develops through a multifactorial process in which various interrelated risk factors contribute to disease susceptibility and progression. Elevated intraocular pressure (IOP) remains the most established modifiable risk factor. , However, patients still experience disease progression despite IOP reduction, and develop glaucoma within the so-called “physiological” IOP range, suggesting that IOP-independent mechanisms also play a significant role in disease development and progression. ,, The vascular theory complements this framework by proposing that impaired blood supply to the optic nerve head contributes to RGC loss, with factors such as systemic hypotension, vasospasm, and atherosclerosis potentially reducing optic nerve perfusion and promoting glaucomatous damage. ,,,,,
More recently, investigators have examined the effects of systemic medications, including β-blockers, calcium channel blockers, and statins, on ocular blood flow and glaucoma-related outcomes. ,,,, Phosphodiesterase-5 inhibitors (PDE-5i) represent a class of systemic agents that enhance endogenous nitric oxide signaling, resulting in smooth muscle relaxation and vasodilation. , Although sildenafil and related agents are primarily prescribed for the treatment of erectile dysfunction (ED) and pulmonary arterial hypertension, growing evidence suggests that their vascular and cytoprotective effects may have broader therapeutic relevance. In Raynaud phenomenon, PDE-5i have been shown to improve digital blood flow and reduce attack frequency and duration, supporting a role in disorders characterized by impaired microvascular perfusion. In cystic fibrosis, experimental and early clinical studies have suggested that PDE-5 inhibition may improve vascular endothelial abnormalities and exercise capacity. In metabolic disease, sildenafil has been associated with improved endothelial function and insulin sensitivity in patients with type 2 diabetes or prediabetes. In neurodegenerative disease, preclinical, physiologic, and observational studies have raised the possibility that sildenafil may enhance cerebral blood flow and modulate pathways relevant to Alzheimer’s disease, although definitive clinical benefit remains unproven.
These vascular effects may also be relevant to ocular diseases. Experimental and clinical studies have shown that PDE-5 inhibition can increase ocular blood flow, particularly within the choroidal circulation, ,,, without consistently affecting IOP. ,,, Because vascular dysregulation and impaired ocular perfusion have been implicated in the pathophysiology of glaucoma, medications that improve microvascular circulation could plausibly influence disease risk. However, the relationship between PDE-5i use and glaucoma remains uncertain. Given the widespread use of these medications and the potential relevance of vascular mechanisms in glaucoma pathogenesis, we conducted this study to evaluate the association between PDE-5i exposure and glaucoma development, hypothesizing that PDE-5i use may be associated with a lower hazard of glaucoma.
MATERIALS AND METHODS
DATA SOURCE
This retrospective cohort study utilized deidentified electronic health record data from the TriNetX United States (U.S.). Collaborative Network, comprising 67 participating health care organizations (HCOs) and more than 130 million patients. Because all data were fully deidentified, the study was considered exempt from Institutional Review Board oversight by the University of Arkansas for Medical Sciences. The study adhered to the Declaration of Helsinki (2013 revision) and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
Clinical conditions were identified using International Classification of Diseases, Tenth Revision (ICD-10) codes, with diagnoses recorded under ICD-9 prior to 2015 internally mapped to their ICD-10 equivalents. Medication exposures were defined using RxNorm identifiers, and procedural data were identified using Current Procedural Terminology (CPT) codes.
INCLUSION/EXCLUSION CRITERIA AND STUDY GROUPS
Male patients aged ≥ 40 years with a diagnosis of ED (N52) were identified and stratified into 2 cohorts: an exposed cohort comprising men who had prescribed sildenafil (RxNorm: 136 411), tadalafil (RxNorm: 358 263), vardenafil (RxNorm: 306 674), or avanafil (RxNorm: 1 291 301), and a comparator group comprising men with ED and no documented PDE-5i use. To restrict the analysis to chronic users, exposure was defined as receipt of at least 4 prescriptions for a PDE-5i agent. Patients with a prior diagnosis of glaucoma suspect (GS) or open-angle glaucoma (OAG) were excluded. Glaucoma suspect and OAG were identified using parent ICD-10 codes H40.0 and H40.1, respectively, which, within the TriNetX platform, encompass all associated subcodes. To enhance outcome ascertainment and minimize detection bias, eligible patients were required to have at least 3 ophthalmology encounters within 3 years following fulfillment of the exposure definition (≥ 4 PDE-5i prescriptions). The index date was defined as the first date on which all inclusion criteria were satisfied and occurred between January 1, 2005, and January 1, 2025. The primary outcomes were the hazards of development of GS and OAG at 1-, 2-, and 3-year intervals following the index date.
STATISTICAL ANALYSIS
Propensity score matching (PSM) was employed to reduce baseline confounding between patients with ED exposed to PDE-5i and those without PDE-5i exposure. A 1:1 nearest-neighbor greedy matching algorithm with a caliper of 0.1 pooled SDs was implemented within the TriNetX platform. Covariate balance after matching was evaluated using standardized mean differences, with values < 0.1 considered indicative of adequate balance.
The matching model incorporated demographic characteristics (age, race, and ethnicity); systemic comorbidities spanning cardiovascular, cerebrovascular, metabolic, renal, pulmonary, neurologic, psychiatric, and substance use disorders; and ophthalmic comorbidities (myopia). Genitourinary conditions and procedures relevant to ED management, including benign prostatic disease, prostate malignancy, penile disorders, and prior medical or surgical interventions for sexual dysfunction, were included to account for disease severity and treatment pathways. Additional covariates comprised key medication classes (antihypertensives, statins, nitrates, testosterone, and incretin-based therapies), laboratory parameters (low-density lipoprotein cholesterol and hemoglobin A1c), biometric measures (body mass index), and socioeconomic indicators related to education, employment, housing, and environmental circumstances.
To further mitigate detection bias related to differences in ophthalmic surveillance, the frequency of ophthalmic encounters and diagnostic testing, including comprehensive and intermediate eye examinations, gonioscopy, retinal nerve fiber layer optical coherence tomography (OCT), and visual field testing, was included in the matching process. The complete list of codes used for covariate construction and matching is provided in E-supplement 1 .
Additionally, Cox proportional hazards models were applied to the unmatched cohorts with adjustment for the same covariates to estimate adjusted hazard ratios (aHRs) and corresponding 95% CIs (CIs). Baseline characteristics were summarized using descriptive statistics, with continuous variables reported as means ± SDs and categorical variables as proportions. All analyses were performed using the TriNetX analytics platform, and statistical significance was defined as a 2-sided P -value <.05.
RESULTS
BASELINE CHARACTERISTICS
Before matching, the cohort included 40,676 males with ED treated with PDE-5i and 32,415 males with ED not treated with PDE-5i. The mean age at index was 61.8 ± 10.2 years in the PDE-5i group compared with 64.3 ± 11.9 years in the no-PDE-5i group. In the PDE-5i cohort, 69.5% were non-Hispanic or Latino (vs 66.7%), 62.2% were White (vs 70.0%), 26.8% were Black or African American (vs 18.7%).
Systemic comorbidities were generally less prevalent in the PDE-5i group, including hypertensive diseases (73.7% vs 75.2%), cerebrovascular disease (11.5% vs 16.3%), chronic kidney disease (14.4% vs 20.0%), and diabetes mellitus (43.7% vs 50.1%). Disorders of lipoprotein metabolism (76.8% vs 73.1%) and sleep disorders (39.8% vs 37.2%) were more frequent among PDE-5i users. Genitourinary and sexual health–related conditions, including testicular dysfunction (13.7% vs 11.4%) and benign prostatic hyperplasia (38.1% vs 36.6%), were more frequent among PDE-5i users. After PSM, both cohorts included 23,603 patients with well-balanced baseline characteristics ( Table 1 ).
TABLE 1
Baseline Characteristics of Erectile Dysfunction Patients Treated With PDE-5i Versus No Exposure, Before and After Propensity Score Matching
| Characteristic | Before Matching | After Matching | ||||
|---|---|---|---|---|---|---|
|
PDE-5i (n, %)
n = 40,676 |
No-PDE-5i (n, %)
n = 32,415 |
Standardized Mean
Difference |
PDE-5i (n, %)
n = 23,603 |
No-PDE-5i (n, %)
n = 23,603 |
Standardized Mean
Difference |
|
| Demographics | ||||||
| Age at index (mean ± SD) | 61.8 ± 10.2 | 64.3 ± 11.9 | 0.227 | 62.9 ± 10.2 | 63.0 ± 11.9 | 0.009 |
| Hispanic or Latino | 2779 (6.8%) | 3343 (10.3%) | 0.125 | 2029 (8.6%) | 2071 (8.8%) | 0.006 |
| Not Hispanic or Latino | 28,243 (69.5%) | 21,563 (66.7%) | 0.061 | 15,803 (67.0%) | 15,816 (67.0%) | 0.001 |
| Unknown ethnicity | 9604 (23.6%) | 7428 (23.0%) | 0.016 | 5771 (24.5%) | 5716 (24.2%) | 0.005 |
| White | 25,265 (62.2%) | 22,643 (70.0%) | 0.166 | 15,968 (67.7%) | 16,004 (67.8%) | 0.003 |
| Black or African American | 10,890 (26.8%) | 6041 (18.7%) | 0.195 | 5030 (21.3%) | 4962 (21.0%) | 0.007 |
| Asian | 1133 (2.8%) | 693 (2.1%) | 0.042 | 574 (2.4%) | 578 (2.4%) | 0.001 |
| American Indian or Alaska Native | 158 (0.4%) | 131 (0.4%) | 0.003 | 105 (0.4%) | 102 (0.4%) | 0.002 |
| Native Hawaiian or other Pacific Islander | 174 (0.4%) | 122 (0.4%) | 0.008 | 102 (0.4%) | 98 (0.4%) | 0.003 |
| Other race | 743 (1.8%) | 1118 (3.5%) | 0.102 | 606 (2.6%) | 620 (2.6%) | 0.004 |
| Unknown race | 2263 (5.6%) | 1586 (4.9%) | 0.03 | 1218 (5.2%) | 1239 (5.2%) | 0.004 |
| Diagnoses | ||||||
| Hypertensive diseases | 29,922 (73.7%) | 24,327 (75.2%) | 0.036 | 17,407 (73.7%) | 17,452 (73.9%) | 0.004 |
| Ischemic heart diseases | 9589 (23.6%) | 10,680 (33.0%) | 0.21 | 6680 (28.3%) | 6760 (28.6%) | 0.008 |
| Cerebrovascular diseases | 4685 (11.5%) | 5258 (16.3%) | 0.137 | 3259 (13.8%) | 3347 (14.2%) | 0.011 |
| Chronic lower respiratory diseases | 10,155 (25.0%) | 8043 (24.9%) | 0.003 | 5913 (25.1%) | 5895 (25.0%) | 0.002 |
| Chronic kidney disease | 5857 (14.4%) | 6471 (20.0%) | 0.149 | 3976 (16.8%) | 3981 (16.9%) | 0.001 |
| Disorders of lipoprotein metabolism | 31,208 (76.8%) | 23,639 (73.1%) | 0.086 | 17,615 (74.6%) | 17,575 (74.5%) | 0.004 |
| Overweight, obesity | 13,984 (34.4%) | 10,954 (33.9%) | 0.011 | 8078 (34.2%) | 8056 (34.1%) | 0.002 |
| Diabetes mellitus | 17,739 (43.7%) | 16,206 (50.1%) | 0.13 | 10,897 (46.2%) | 10,935 (46.3%) | 0.003 |
| Sleep disorders | 16,149 (39.8%) | 12,023 (37.2%) | 0.053 | 9004 (38.1%) | 8970 (38.0%) | 0.003 |
| Disorders of thyroid gland | 5370 (13.2%) | 4760 (14.7%) | 0.043 | 3304 (14.0%) | 3306 (14.0%) | 0.001 |
| Other hypothyroidism | 3660 (9.0%) | 3471 (10.7%) | 0.058 | 2336 (9.9%) | 2360 (10.0%) | 0.003 |
| Thyrotoxicosis | 743 (1.8%) | 575 (1.8%) | 0.004 | 415 (1.8%) | 415 (1.8%) | 0.001 |
| Nicotine dependence | 8161 (20.1%) | 6132 (19.0%) | 0.028 | 4631 (19.6%) | 4665 (19.8%) | 0.004 |
| Alcohol related disorders | 4114 (10.1%) | 2967 (9.2%) | 0.032 | 2263 (9.6%) | 2298 (9.7%) | 0.005 |
| Opioid related disorders | 1167 (2.9%) | 941 (2.9%) | 0.002 | 659 (2.8%) | 684 (2.9%) | 0.006 |
| Cocaine related disorders | 1005 (2.5%) | 636 (2.0%) | 0.034 | 508 (2.2%) | 510 (2.2%) | 0.001 |
| Mental & behavioral disorders due to substance use | 11,415 (28.1%) | 8502 (26.3%) | 0.041 | 6410 (27.2%) | 6475 (27.4%) | 0.006 |
| Depressive episode | 9051 (22.3%) | 7567 (23.4%) | 0.027 | 5410 (22.9%) | 5408 (22.9%) | 0.001 |
| Other anxiety disorders | 9139 (22.5%) | 6816 (21.1%) | 0.034 | 5213 (22.1%) | 5201 (22.0%) | 0.001 |
| Multiple sclerosis | 326 (0.8%) | 215 (0.7%) | 0.016 | 177 (0.7%) | 165 (0.7%) | 0.006 |
| Parkinson disease | 387 (1.0%) | 545 (1.7%) | 0.064 | 294 (1.2%) | 282 (1.2%) | 0.005 |
| Paraplegia/quadriplegia | 187 (0.5%) | 213 (0.7%) | 0.027 | 133 (0.6%) | 132 (0.6%) | 0.001 |
| Myopia | 7830 (19.3%) | 6033 (18.7%) | 0.016 | 4696 (19.9%) | 4640 (19.7%) | 0.006 |
| Metabolic syndrome | 907 (2.2%) | 662 (2.0%) | 0.013 | 509 (2.2%) | 515 (2.2%) | 0.002 |
| Testicular dysfunction | 5559 (13.7%) | 3695 (11.4%) | 0.068 | 2880 (12.2%) | 2849 (12.1%) | 0.004 |
| Hyperfunction of pituitary gland | 398 (1.0%) | 321 (1.0%) | 0.001 | 223 (0.9%) | 227 (1.0%) | 0.002 |
| Benign prostatic hyperplasia | 15,472 (38.1%) | 11,823 (36.6%) | 0.031 | 8698 (36.9%) | 8786 (37.2%) | 0.008 |
| Other disorders of penis | 2866 (7.1%) | 2142 (6.6%) | 0.017 | 1539 (6.5%) | 1555 (6.6%) | 0.003 |
| Inflammatory diseases of prostate | 2695 (6.6%) | 1428 (4.4%) | 0.097 | 1191 (5.0%) | 1202 (5.1%) | 0.002 |
| Acquired absence of genital organs | 1001 (2.5%) | 1177 (3.6%) | 0.068 | 695 (2.9%) | 712 (3.0%) | 0.004 |
| Personal history of prostate cancer | 1670 (4.1%) | 2647 (8.2%) | 0.17 | 1378 (5.8%) | 1388 (5.9%) | 0.002 |
| Procedures | ||||||
| Cataract surgery– standard | 2166 (5.3%) | 1659 (5.1%) | 0.009 | 1250 (5.3%) | 1223 (5.2%) | 0.005 |
| Cataract surgery– complex | 503 (1.2%) | 468 (1.4%) | 0.018 | 329 (1.4%) | 320 (1.4%) | 0.003 |
| Introduction procedures on penis | 1003 (2.5%) | 818 (2.5%) | 0.004 | 519 (2.2%) | 519 (2.2%) | 0.001 |
| Surgical procedures on penis | 1466 (3.6%) | 1497 (4.6%) | 0.051 | 775 (3.3%) | 768 (3.3%) | 0.002 |
| Injection– corpora cavernosa | 931 (2.3%) | 775 (2.4%) | 0.007 | 491 (2.1%) | 481 (2.0%) | 0.003 |
| Irrigation– priapism | 44 (0.1%) | 37 (0.1%) | 0.002 | 24 (0.1%) | 26 (0.1%) | 0.003 |
| Injection– alprostadil | 52 (0.1%) | 46 (0.1%) | 0.004 | 21 (0.1%) | 25 (0.1%) | 0.005 |
| Injection– papaverine | 105 (0.3%) | 107 (0.3%) | 0.013 | 66 (0.3%) | 67 (0.3%) | 0.001 |
| Injection– phentolamine | 13 (0.0%) | 30 (0.1%) | 0.024 | 11 (0.0%) | 10 (0.0%) | 0.002 |
| Injection– testosterone | 331 (0.8%) | 114 (0.4%) | 0.061 | 101 (0.4%) | 108 (0.5%) | 0.004 |
| Therapeutic injection | 8043 (19.8%) | 6042 (18.7%) | 0.028 | 4582 (19.4%) | 4595 (19.5%) | 0.001 |
| Inflatable penile prosthesis | 63 (0.2%) | 387 (1.2%) | 0.127 | 61 (0.3%) | 56 (0.2%) | 0.004 |
| Inflatable device (supply) | 33 (0.1%) | 215 (0.7%) | 0.096 | 32 (0.1%) | 32 (0.1%) | 0.001 |
| Prosthetic implant NOS | 350 (0.9%) | 253 (0.8%) | 0.009 | 191 (0.8%) | 198 (0.8%) | 0.003 |
| Medications | ||||||
| Corticosteroids | 28,026 (69.0%) | 19,194 (59.4%) | 0.202 | 15,095 (64.0%) | 15,038 (63.7%) | 0.005 |
| Statins | 27,343 (67.3%) | 18,997 (58.8%) | 0.178 | 14,900 (63.1%) | 14,879 (63.0%) | 0.002 |
| Angiotensin-converting enzyme (ACE) inhibitors | 18,228 (44.9%) | 12,715 (39.3%) | 0.112 | 9863 (41.8%) | 9969 (42.2%) | 0.009 |
| Diuretics | 17,900 (44.1%) | 13,139 (40.6%) | 0.069 | 9933 (42.1%) | 9968 (42.2%) | 0.003 |
| Beta blockers | 17,474 (43.0%) | 14,825 (45.8%) | 0.057 | 10,556 (44.7%) | 10,646 (45.1%) | 0.008 |
| Calcium channel blockers | 14,748 (36.3%) | 10,624 (32.9%) | 0.072 | 8106 (34.3%) | 8103 (34.3%) | 0.001 |
| Angiotensin II inhibitors | 9637 (23.7%) | 6743 (20.9%) | 0.069 | 5293 (22.4%) | 5291 (22.4%) | 0.001 |
| Nitroglycerin | 5381 (13.2%) | 5805 (18.0%) | 0.13 | 3783 (16.0%) | 3774 (16.0%) | 0.001 |
| Isosorbide | 862 (2.1%) | 2062 (6.4%) | 0.212 | 811 (3.4%) | 797 (3.4%) | 0.003 |
| Isosorbide dinitrate | 275 (0.7%) | 657 (2.0%) | 0.117 | 243 (1.0%) | 244 (1.0%) | 0.001 |
| Clonidine | 1766 (4.3%) | 1554 (4.8%) | 0.022 | 1101 (4.7%) | 1083 (4.6%) | 0.004 |
| Dexmedetomidine | 1680 (4.1%) | 1817 (5.6%) | 0.069 | 1183 (5.0%) | 1200 (5.1%) | 0.003 |
| Testosterone | 3976 (9.8%) | 1848 (5.7%) | 0.153 | 1685 (7.1%) | 1629 (6.9%) | 0.009 |
| Insulin | 9828 (24.2%) | 9137 (28.3%) | 0.093 | 6250 (26.5%) | 6248 (26.5%) | 0.001 |
| Metformin | 13,031 (32.1%) | 8991 (27.8%) | 0.093 | 7141 (30.3%) | 7220 (30.6%) | 0.007 |
| Oral hypoglycemic agents | 14,358 (35.3%) | 10,702 (33.1%) | 0.047 | 8085 (34.3%) | 8178 (34.6%) | 0.008 |
| Sitagliptin | 2600 (6.4%) | 1638 (5.1%) | 0.057 | 1338 (5.7%) | 1329 (5.6%) | 0.002 |
| Empagliflozin | 2129 (5.2%) | 1600 (4.9%) | 0.013 | 1219 (5.2%) | 1254 (5.3%) | 0.007 |
| Pioglitazone | 1739 (4.3%) | 1208 (3.7%) | 0.028 | 931 (3.9%) | 930 (3.9%) | 0.001 |
| Glyburide | 1524 (3.8%) | 992 (3.1%) | 0.038 | 762 (3.2%) | 789 (3.3%) | 0.006 |
| Glimepiride | 1670 (4.1%) | 1313 (4.1%) | 0.003 | 943 (4.0%) | 976 (4.1%) | 0.007 |
| Dapagliflozin | 784 (1.9%) | 658 (2.0%) | 0.008 | 510 (2.2%) | 492 (2.1%) | 0.005 |
| Canagliflozin | 729 (1.8%) | 407 (1.3%) | 0.044 | 378 (1.6%) | 349 (1.5%) | 0.01 |
| Linagliptin | 518 (1.3%) | 408 (1.3%) | 0.001 | 281 (1.2%) | 295 (1.2%) | 0.005 |
| Saxagliptin | 181 (0.4%) | 92 (0.3%) | 0.027 | 82 (0.3%) | 77 (0.3%) | 0.004 |
| Alogliptin | 111 (0.3%) | 46 (0.1%) | 0.029 | 46 (0.2%) | 39 (0.2%) | 0.007 |
| Other hypoglycemic agents | 3727 (9.2%) | 2452 (7.6%) | 0.057 | 2006 (8.5%) | 1982 (8.4%) | 0.004 |
| Dulaglutide | 1643 (4.0%) | 1075 (3.3%) | 0.038 | 869 (3.7%) | 866 (3.7%) | 0.001 |
| Semaglutide | 1592 (3.9%) | 1004 (3.1%) | 0.044 | 854 (3.6%) | 827 (3.5%) | 0.006 |
| Liraglutide | 1185 (2.9%) | 718 (2.2%) | 0.044 | 595 (2.5%) | 588 (2.5%) | 0.002 |
| Tirzepatide | 491 (1.2%) | 304 (0.9%) | 0.026 | 270 (1.1%) | 258 (1.1%) | 0.005 |
| Exenatide | 595 (1.5%) | 301 (0.9%) | 0.049 | 252 (1.1%) | 245 (1.0%) | 0.003 |
| Glipizide | 4166 (10.3%) | 3055 (9.4%) | 0.027 | 2322 (9.8%) | 2318 (9.8%) | 0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
