Abstract
This study evaluated technicians’ ability to identify sperm morphology via external quality assessment (EQA) and identified challenging defect types causing significant interparticipant discrepancies. A longitudinal study was conducted from 2019 to 2024 to assess the ability of technicians at various laboratories to identify sperm morphology using standardized EQA protocols and distributed sperm images. Morphological classification was performed using the criteria of the World Health Organization, 6th edition. Correct identification rates, participant pass rates, and defect-specific accuracy rates were statistically analyzed. The overall correct identification rate (mean ± standard deviation [s.d.]) was 81.3% ± 27.0%, while the correct identification rate of normal sperm morphology (mean ± s.d.) was only 51.1% ± 25.5%. Although the overall correct identification rate of abnormal sperm morphology (mean ± s.d.) was as high as 96.1% ± 8.9%, the overall correct identification rate across all defect categories (mean ± s.d.) was only 60.2% ± 21.7%. The identification of sperm head defects was the most challenging, with a correct identification rate (mean ± s.d.) of only 57.0% ± 20.7%, while the correct identification rate of sperm tail defects (mean ± s.d.) was the highest at 70.5% ± 23.5%. Among sperm head defects, the correct identification rate of head size defects (mean ± s.d.) was the lowest at 53.4% ± 22.0%. Laboratories that participated for 5 consecutive years showed significant improvements in total and normal morphology scores. Image-based EQA effectively assesses the technicians’ competency and serves as an educational tool. Normal sperm morphology evaluation should avoid overly strict criteria, while abnormal sperm morphology assessment requires enhanced focus on specific defect types.
Keywords: defect identification, external quality assessment, male infertility, sperm morphology, technicians’, ability
INTRODUCTION
Sperm morphology has long been a key diagnostic tool in male infertility that correlates with the outcomes of natural conception, intrauterine insemination, in vitro fertilization, and intracytoplasmic sperm injection.1,2,3,4,5,6 However, since the 5th edition of the World Health Organization (WHO) manual for the examination and processing of human semen was released in 20107 and recommended the use of strict criteria to assess sperm morphology, studies have reported that the clinical usefulness of sperm morphology assessments has decreased.7,8,9,10,11,12,13 Notably, men without normal sperm based on strict criteria can still achieve pregnancy without in vitro fertilization, challenging the value of sperm morphology as a predictor of fertilization success.14
The discrepant research findings and lack of a definitive consensus on the impact of sperm morphology on reproductive outcomes may be attributed to multiple factors. One of the main reasons is the significant differences in morphological results within and between laboratories.15 The manual assessment of sperm morphology is highly subjective. Although andrology laboratories conduct internal and external training, there are still large variations in sperm morphology results, and the highest coefficient of variation reported in the literature is 375.0%.16 This highly significant variation calls into question the reproducibility of any sperm morphology assessment, even when using similar samples. The definition and interpretation of morphological criteria may be another important reason leading to the loss of value of sperm morphology assessment. The criteria have become increasingly stringent over time, resulting in a corresponding decline in the average number of normal spermatic forms per year.10
The 6th edition of the WHO manual17 states that individuals must be familiar with all relevant criteria to conduct a proper sperm morphology assessment. This implies that even if a laboratory opts to merely report the proportions of normal or abnormal spermatozoa, both internal quality control and external quality assessment (EQA) are needed to evaluate the ability of technicians to accurately identify abnormalities in all locations. However, there is a lack of research on the use of EQA to evaluate technicians’ ability to accurately identify sperm morphological abnormalities. In this study, we aimed to evaluate the ability of technicians to correctly recognize sperm morphology through EQA programs. Additionally, we aimed to determine which specific sperm morphology defects were not easily identified accurately, as indicated by significant differences among the participants.
PATIENTS AND METHODS
Scheme organization
The EQA program was organized by the Andrology Laboratory/Human Sperm Bank of Sichuan (West China Second University Hospital, Sichuan University [WCSUH-SCU], Chengdu, China), an ISO 9001-certified laboratory. Ethical approval was obtained from the Ethics Review Board of WCSUH-SCU (Approval No. 2023-072). All participants provided written informed consent that included permission for the use of anonymized data for research publication. The EQA programs were carried out from 2019 to 2024, except in 2020. The assessed items were semen pH, sperm concentration, sperm motility, sperm morphology, sperm viability, DNA fragmentation index, high DNA stainability, sperm chromatin (aniline blue staining condition), and seminal plasma biochemical (zinc, citric acid, fructose, and neutral α-glucosidase).
For the assessment of sperm morphology, the participants were sent two unstained slides and images of sperm cells. The WCSUH-SCU prepared semen smears and then carried out modified Papanicolaou staining. Images were collected using the automated sperm morphology analysis (ASMA) system (Puhua Life, Chengdu, China) with an oil immersion magnification of 1000×, and a sperm morphology image library was formed. The ASMA system measures 14 parameters of the sperm head and assists technicians in classifying sperm morphology.18 Both the original image and the magnified image were provided for each sperm. The original image enabled the evaluators to assess the actual size of the sperm, while the magnified image allowed them to evaluate the defects of each part.
According to the sperm morphology classification requirements in the 5th and 6th edition of the WHO manual,7,17 sperms with abnormal forms were divided into four major categories and 24 subcategories: head defects (macrocephalic, microcephalic, pyriform, tapered, round, non-oval, amorphous, more than two small vacuoles, large vacuoles, post-acrosomal vacuoles, small acrosome, large acrosome, acephalic, and multiple heads), neck/midpiece defects (bent neck, asymmetrical, thick insertion, and thin), tail defects (short, bent, coiled [broken], tailless sperm, and multiple tails), and excessive residual cytoplasm (ERC).
The organizers selected and distributed images of sperm with normal or abnormal morphology from the image library annually. A total of 450 sperm images were distributed from 2019 to 2024.
Statistical analyses
For the EQA results, data with normal distribution were presented as mean ± standard deviation (s.d.), whereas data with non-normal distribution were presented as the median and 25th and 75th percentiles (P25, P75). Between-group comparisons were performed using Wilcoxon’s rank-sum test. Statistical analysis was performed using SPSS version 25.0 (IBM Corporation, Armonk, NY, USA). P < 0.05 was considered statistically significant. Pass rate (%) is defined as: (number of laboratories with a score ≥80)/(total number of laboratories) × 100%. Correct rate (%) is defined as: (number of participants who correctly identify sperm morphology)/(number of participants in that year) × 100%. The “correct rate” was defined as the percentage of participants who accurately identified a given sperm morphology feature compared with the consensus-based gold standard classification established by a panel of five experienced andrology technicians. Each panel member had a minimum of 10 years of experience in sperm morphology assessment, (co)authored peer-reviewed publications on the topic,18 and had extensive experience in teaching national training courses on semen analysis. The members of the panel each independently assessed each image and then convened to review all images for which there were initial disagreements. A final consensus classification (normal/abnormal and specific defect type) was reached for every sperm image through detailed discussion, aided by the morphometric parameters provided by the ASMA system.
RESULTS
Participant information
From 2019 to 2024, the number of participants increased annually. In 2024, 191 laboratories from 94 cities in 29 provinces in China participated in the EQA program for sperm morphology (Figure 1). The laboratories participating in EQA were mainly andrology laboratories, and the participating hospitals were predominantly tertiary level. Most laboratories had full-time technicians who had received relevant training (Table 1).
Figure 1.
Numbers of participants in the external quality assessment program for sperm morphology from 2019 to 2024.
Table 1.
Basic information about laboratories participating in the external quality assessment program in 2024
| Basic information | Number of laboratories (total number of participants=191), n (%) |
|---|---|
| Type of laboratories involved | |
| Andrology laboratory | 130 (68.1) |
| Clinical laboratory | 48 (25.1) |
| Human sperm bank | 9 (4.7) |
| Others | 4 (2.1) |
| Hospital grade | |
| Tertiary hospital | 151 (79.1) |
| Secondary hospital | 14 (7.3) |
| Primary hospital | 26 (13.6) |
| Types of technicians involved | |
| At least one full-time technician | 160 (83.8) |
| Only part-time technicians | 31 (16.2) |
| Training undertaken | |
| At least one technician has been trained | 162 (84.8) |
| None of the technicians have been trained | 29 (15.2) |
EQA scores and pass rates
The total EQA scores, normal sperm morphology scores, abnormal sperm morphology scores, and pass rates of all participants are shown in Table 2. From 2019 to 2024, the median score (P25, P75) for abnormal sperm morphology was 100.0 (97.6, 100.0), which was higher than that for normal sperm morphology of 63.0 (44.0, 81.0). During the 5-year study period, there were significant differences between normal and abnormal sperm morphology scores (Wilcoxon’s rank-sum test, P < 0.05), which suggests that the participants were overly strict in distinguishing the morphology of normal sperm. Therefore, the number of normal sperm images was increased among the 100 sperm images distributed in 2024. This may explain why the overall scores and pass rates of participating laboratories increased from 2019 to 2023 but declined in 2024.
Table 2.
External quality assessment results for 450 sperm images (2019–2024)
| Year | Number of sperm images (n) | Total score | Number of normal sperm morphology images (n) | Normal sperm morphology score | Number of abnormal sperm morphology images (n) | Abnormal sperm morphology score | Pass rate (%) |
|---|---|---|---|---|---|---|---|
| 2019 | 100 | 53.0 (51.0–61.0) | 58 | 21.6 (16.8–34.9) | 42 | 98.8 (97.0–100.0)* | 14.3 |
| 2021 | 100 | 88.0 (84.0–90.0) | 18 | 64.6±21.2 | 82 | 95.1 (86.6–98.8)* | 88.9 |
| 2022 | 50 | 90.0 (86.0–92.0) | 8 | 50.0 (37.5–62.5) | 42 | 97.6 (95.2–100.0)* | 96.9 |
| 2023 | 100 | 96.0 (94.0–98.0) | 14 | 78.6 (60.7–92.9) | 86 | 100.0 (98.8–100.0)* | 100.0 |
| 2024 | 100 | 81.0 (73.5–89.0) | 48 | 63.0 (45.0–79.0) | 52 | 100.0 (98.0–100.0)* | 56.5 |
| Total | 450 | 89.0 (79.0–94.0) | 146 | 63.0 (44.0–81.0) | 304 | 100.0 (97.6–100.0)* | 71.32 |
*P<0.05, Wilcoxon’s rank-sum test, comparison of scores between normal sperm morphology and abnormal sperm morphology. Normally distributed data are presented as mean±s.d., whereas non-normally distributed data are presented as median (25th percentile, 75th percentile). Pass rate (%)=(number of laboratories with a score ≥80)/(total number of laboratories) ×100%. s.d.: standard deviation
Results of the 10 participating laboratories in 2019 and 2024
Among the 14 participating laboratories in 2019, 10 participated in the EQA for 5 consecutive years. The total median (P25, P75) scores of eight of these 10 laboratories increased from 53.0 (49.8, 68.3) in 2019 to 86.5 (78.0, 96.0) in 2024 (Figure 2a). The normal sperm morphology median (P25, P75) scores of seven laboratories increased from 21.6 (16.4, 47.0) in 2019 to 73.0 (54.0, 98.0) in 2024 (Figure 2b). However, there was little change in the abnormal sperm morphology scores from 2019 to 2024 (Figure 2c).
Figure 2.
Comparison of (a) the total scores, (b) normal sperm scores, and (c) abnormal sperm scores of 10 laboratories between 2019 and 2024.
Correct rate of participants’ evaluation of sperm morphology
The correct rates of all sperm, normal sperm morphology, abnormal sperm morphology, and 24 types of sperm defects are shown in Supplementary Table 1. Over the 5-year study period, the overall correct rate for the evaluation of 450 sperm images was relatively high at 81.3% ± 27.0%. The correct rate showed a steady increase from 2019 (58.8% ± 36.5%) to 2023 (96.2% ± 11.9%), followed by a slight decline in 2024 (80.6% ± 23.7%). The cause of the decline in 2024 might be owing to either the increase in the number of participants in 2024 or the increase in the number of distributed normal sperm images (Figure 3a).
Supplementary Table 1.
Correct rate of participant's evaluation of sperm morphology on 450 images of sperm
| Parameters | 2019 | 2021 | 2022 | 2023 | 2024 | Total | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
|||||||
| Number of images | Correct rate (n/N, %) | Number of images | Correct rate (n/N, %) | Number of images | Correct rate (n/N, %) | Number of images | Correct rate (n/N, %) | Number of images | Correct rate (n/N, %) | Number of images | Correct rate (n/N, %) | |
| All sperm | 100 | 58.8±36.5 | 100 | 85.1±19.0 | 50 | 90.4±16.3 | 100 | 96.2±11.9 | 100 | 80.6±23.7 | 450 | 81.3±27.0 |
| Normal sperm morphology | 58 | 31.0±19.6 | 18 | 61.7±25.7 | 8 | 61.3±19.2 | 14 | 70.3±18.4 | 48 | 64.1±22.2 | 146 | 51.1±25.5 |
| Abnormal sperm morphology | 42 | 97.0±9.3 | 82 | 90.2±12.5 | 42 | 95.9±7.6 | 86 | 99.8±1.0 | 52 | 98.9±4.1 | 304 | 96.1±8.9 |
| Abnormal categories | 57 | 39.6±20.5 | 125 | 52.6±22.5 | 66 | 60.4±23.2 | 173 | 68.1±18.7 | 86 | 68.4±19.5 | 507 | 60.2±21.7 |
| Head defects | 51 | 36.2±12.3 | 99 | 50.8±20.6 | 45 | 61.5±21.1 | 117 | 64.2±18.5 | 66 | 66.7±17.6 | 378 | 57.0±20.7 |
| Macrocephalic | 0 | / | 0 | / | 0 | / | 8 | 35.2±24.3 | 0 | / | 8 | 35.2±24.323 |
| Microcephalic | 1 | 21.3 | 5 | 46.2±10.2 | 3 | 59.9±15.2 | 6 | 67.2±15.7 | 5 | 43.7±2.1 | 20 | 52.7±16.818 |
| Pyriform | 3 | 45.2±2.1 | 22 | 49.8±16.4 | 8 | 59.0±15.6 | 17 | 66.2±14.6 | 16 | 75.6±13.0 | 66 | 61.2±20.015 |
| Tapered | 3 | 38.1±2.1 | 13 | 56.8±15.7 | 4 | 75.4±13.1 | 10 | 72.9±14.9 | 11 | 72.8±8.2 | 41 | 65.5±15.713 |
| Round | 2 | 48.2±2.5 | 3 | 73.1±16.3 | 2 | 57.9±39.1 | 9 | 81.2±5.1 | 5 | 80.6±9.8 | 21 | 74.5±17.68 |
| Non-oval | 11 | 33.4±10.4 | 15 | 35.4±20.8 | 8 | 57.2±15.8 | 15 | 50.2±20.1 | 0 | / | 49 | 43.0±22.621 |
| Amorphous | 3 | 46.4±0 | 2 | 65.4±3.9 | 4 | 78.9±10.1 | 5 | 72.0±13.9 | 2 | 39.8±8.9 | 16 | 64.1±18.314 |
| More than two small vacuoles | 3 | 40.5±16.8 | 7 | 54.4±25.2 | 2 | 82.1±2.4 | 6 | 79.1±17.8 | 2 | 92.7±1.1 | 20 | 66.3±25.612 |
| Large vacuoles | 1 | 85.7 | 0 | / | 0 | / | 0 | / | 0 | / | 1 | 85.72 |
| Post-acrosomal vacuoles | 4 | 31.3±12.2 | 5 | 49.2±19.9 | 3 | 74.0±9.0 | 4 | 56.1±6.1 | 4 | 61.0±22.7 | 20 | 53.1±19.317 |
| Small acrosomal | 20 | 32.0±7.6 | 22 | 57.2±21.4 | 6 | 57.0±21.4 | 29 | 67.5±10.6 | 15 | 63.3±11.4 | 92 | 56.0±15.716 |
| Large acrosomal | 0 | / | 4 | 38.5±7.2 | 2 | 18.8±3.2 | 8 | 59.4±15.1 | 6 | 52.4±21.6 | 20 | 49.1±19.719 |
| Acephalic | 0 | / | 0 | / | 1 | 90.6 | 0 | / | 0 | / | 1 | 90.61 |
| Multiple heads | 0 | / | 1 | 42.3 | 2 | 33.6±13.3 | 0 | / | 0 | / | 3 | 36.5±11.622 |
| Neck and midpiece defects | 3 | 57.1±25.8 | 13 | 61.1±34.3 | 11 | 57.5±30.8 | 19 | 75.4±17.4 | 6 | 83.8±9.9 | 52 | 70.0±26.0 |
| Bent neck | 1 | 64.3 | 6 | 80.8±17.9 | 5 | 59.7±27.7 | 11 | 76.3±20.8 | 5 | 86.2±8.9 | 28 | 75.6±20.97 |
| Asymmetrical | 1 | 28.5 | 4 | 24.6±17.3 | 1 | 10.9 | 0 | / | 0 | / | 6 | 23.0±16.6 |
| Thick insertion | 1 | 78.6 | 3 | 70.5±17.8 | 4 | 76.2±18.2 | 8 | 74.1±12.3 | 1 | 71.7 | 17 | 74.1±14.69 |
| Thin | 0 | / | 0 | / | 1 | 18.8 | 0 | / | 0 | / | 1 | 18.823 |
| Tail defects | 2 | 96.5±3.6 | 4 | 61.6±34.0 | 8 | 59.8±25.0 | 28 | 75.7±17.9 | 10 | 62.4±29.5 | 52 | 70.5±23.5 |
| Short | 1 | 92.9 | 2 | 32.7±8.2 | 2 | 32.9±18.8 | 5 | 56.5±8.1 | 4 | 33.1±12.8 | 14 | 46.2±21.920 |
| Bent | 0 | / | 0 | / | 0 | / | 1 | 79.6 | 0 | / | 1 | 79.65 |
| Coiled (broken) | 0 | / | 2 | 90.4±5.8 | 3 | 62.0±21.3 | 14 | 79.4±10.7 | 5 | 80.7±18.5 | 24 | 78.4±20.06 |
| Tailless | 0 | / | 0 | / | 1 | 84.4 | 8 | 80.6±12.4 | 1 | 88.5 | 10 | 81.8±11.03 |
| Multiple tails | 1 | 100.0 | 0 | / | 2 | 71.1±0.8 | 0 | / | 0 | / | 3 | 80.7±13.64 |
| Excess residual cytoplasm | 1 | 47.6 | 9 | 55.6±17.5 | 2 | 54.0±2.4 | 9 | 80.6±11.3 | 4 | 88.0±6.6 | 25 | 69.3±19.210 |
Each sperm cell was classified into specific defect types according to the criteria stipulated in the 5th edition (2010) and the 6th edition (2021) of the World Health Organization. n: the number of participants who correctly identify sperm morphology; N: the number of participants in that year. The number of participants in 2019, 2021, 2022, 2023, and 2024 was 14, 27, 64, 103, and 191, respectively. Correct rate=n ÷ N × 100;/: no such defect this year. If the quantity is only one, the data presented in the table is the original data. If the quantity is ≥2, the data shall be presented as mean±s.d.1-24 Ranking of the correct rate of 24 defects from highest to lowest. s.d.: standard deviation
Figure 3.
Correct identification rates for sperm morphology categories across EQA cycles (2019–2024). (a) Correct rates of all sperm, normal sperm morphology, abnormal sperm morphology, and abnormal categories. (b) Correct rates of head defects, neck and midpiece defects, tail defects, and excess residual cytoplasm defects. Correct rate (%) = (number of participants who correctly identified sperm morphology)/(number of participants in that year) × 100%.
Among the 450 sperm images, 146 were classified as having normal sperm morphology, with a correct rate of 51.1% ± 25.5%. This low correct rate indicates that participants were overly strict in their evaluation of normal sperm morphology, leading to a high rate of misclassification. The correct rate for normal sperm morphology gradually increased from 2019 (31.0% ± 19.6%) to 2023 (70.3% ± 18.4%) but decreased slightly in 2024 (64.1% ± 22.2%). In contrast, the correct rate for abnormal sperm morphology was significantly higher, with a correct rate of 96.1% ± 8.9% achieved for 304 abnormal sperm images. This suggests that participants were more accurate in identifying abnormal sperm compared with normal sperm (Figure 3a).
As each sperm could have one or multiple defects, the same sperm could have a maximum of four defect categories. The correct rate for identifying defect categories steadily improved from 39.6% ± 20.5% in 2019 to 68.4% ± 19.5% in 2024. However, the overall correct rate for defect classification remained only 60.2% ± 21.7%, which was significantly lower than the correct rate of 96.1% for abnormal sperm morphology. This indicates that while technicians correctly recognized abnormal sperm morphology, they often struggled to accurately classify specific defect types (Figure 3a).
The overall correct rates for head defects, neck and midpiece defects, tail defects, and ERC were 57.0% ± 20.7%, 70.0% ± 26.0%, 70.5% ± 23.5%, and 69.3% ± 19.2%, respectively. The correct rate was the lowest for head defects and the highest for tail defects. The correct rates for head defects, neck and midpiece defects, and ERC all significantly increased over time from 36.2% ± 12.3%, 57.1% ± 25.8%, and 47.6%, respectively, in 2019 to 66.7% ± 17.6%, 83.8% ± 9.9%, and 88.0% ± 6.6%, respectively, in 2024 (Figure 3b).
For the 24 specific defect categories (Supplementary Table 1), the overall correct rate rankings for 2019–2024 from highest to lowest were as follows: acephalic (90.6%), large vacuoles (85.7%), tailless (81.8% ± 11.0%), multiple tails (80.7% ± 13.6%), bent tails (79.6%), coiled (broken) tails (78.4% ± 20.0%), bent neck (75.6% ± 20.9%), round heads (74.5% ± 17.6%), thick insertion for the mid-piece (74.1% ± 14.6%), ERC (69.3% ± 19.2%), more than two small vacuoles (66.3% ± 25.6%), tapered heads (65.5% ± 15.7%), amorphous heads (64.1% ± 18.3%), pyriform heads (61.2% ± 20.0%), small acrosome (56.0% ± 15.7%), post-acrosomal vacuoles (53.1% ± 19.3%), microcephalic (52.7% ± 16.8%), large acrosomal heads (49.1% ± 19.7%), short tails (46.2% ± 21.9%), non-oval heads (43.0% ± 22.6%), multiple heads (36.5% ± 11.6%), macrocephalic (35.2% ± 24.3%), asymmetrical neck (23.0% ± 16.6%), and thin neck (18.8%).
The sperm head defects were classified into size defects (Figure 4a), shape defects (Figure 4b), and structure defects (Figure 4c), with correct rates of 53.4% ± 22.0%, 59.2% ± 17.8%, and 59.4% ± 24.1%, respectively. The overall correct rate of head size defects was the lowest among the three types of head defects. The overall correct rate of head defects increased over time. In contrast, the correct rates for neck and midpiece defects and various tail defects remained stable or slightly decreased (Figure 4d and 4e).
Figure 4.
Overall correct rates in the 23 morphological specific defect categories (correct rate of excess residual cytoplasm is shown in Figure 3). (a) Correct rate of head size defects (macrocephalic, microcephalic, large acrosome, and small acrosome). (b) Correct rate of head shape defects (pyriform, tapered, round, non-oval, and amorphous). (c) Correct rate of head structural defects (more than two small vacuoles, large vacuoles, post-acrosomal vacuoles, acephalic, and multiple heads). (d) Correct rates of neck defects (bent neck, asymmetrical, thick insertion, and thin neck). (e) Correct rates of tail defects (short, bent, coiled [broken], tailless, and multiple tails). Correct rate (%) = (number of participants who correctly identified sperm morphology)/(number of participants in that year) × 100%.
DISCUSSION
This study revealed a low overall correct identification rate for normal sperm morphology of 51.1% ± 25.5%. This result is far below the expected level, suggesting that technicians may apply excessively stringent criteria during the sperm morphology assessment. Additionally, the organizers surveyed the data of 191 male infertility clinic patients in 2023. Among them, 94 participants provided the average percentage of normal sperm morphology in that year. The median of the 94 annual averages was 5.7% (unpublished data), which is significantly lower than the 50th percentile (14%) of the normal fertile population in the 6th edition of the WHO manual17 and even lower than the 25th percentile (8%). These clinical findings further support the hypothesis that participants were overly strict in their sperm morphology assessment, which may lead to an overdiagnosis of teratozoospermia. Moreover, feedback from clinicians showed that most patients diagnosed with teratozoospermia had a history of achieving pregnancy or giving birth, making it difficult to accurately identify patients with infertility caused by teratozoospermia. These findings underscore the need to recalibrate “normal” sperm morphology criteria and prioritize the identification of clinically relevant abnormalities to restore diagnostic utility in infertility management.19
The correct rate for the identification of abnormal sperm morphology in specific defect categories was only 60.2% ± 21.7%, which was significantly lower than the overall accuracy rate for identifying abnormal sperm (96.1% ± 8.9%). This result highlights the persistent challenges in accurately pinpointing the types of sperm defects, indicating that some defect types that should be identified are not correctly recognized, while some defect types that are not present are mistakenly identified.
Sperm head defects were the most challenging abnormalities to identify, with a correct rate of 57.0% ± 20.7%. In particular, although the correct rate for the identification of head size defects increased over time, the overall rate was still low. This phenomenon may be because technicians lack sufficient objectivity when they manually identify the size of sperm, especially those with critical sizes, compared with instrumental identification. Among the head shape defects, the correct identification rate of non-oval heads was the lowest at only 43.0% ± 22.6%. The main reason may be that different technicians have different understandings of the “generally oval” standard. Some technicians adopt strict criteria, while others are relatively lenient. Clinical data from male infertility clinic patients show a low annual average percentage of normal sperm morphology. This finding indicates that when assessing sperm head shape, technicians might relax the oval standard, accepting a “generally oval” shape. In addition, there were also difficulties in the identification of head structure defects, especially for the relatively rare defect types, such as multiheaded sperm (with a correct identification rate of 36.5% ± 11.6%). When a sperm has multiple heads or a head contains multiple nuclei, technicians may misjudge it as overlapping sperm or immature sperm cells. Wang et al.15 reached a similar conclusion that the assessment of sperm heads is more challenging for experts than the assessment of other parts of sperm cells. Given the high abnormal rate of sperm head defects in clinical practice, the accurate identification of sperm head defects is particularly important.
The organizers held a summary meeting every year to provide participants with feedback on the results of the EQA and conduct relevant training on sperm morphology assessment. These measures likely contributed to the marked improvement among the 10 laboratories that participated for 5 consecutive years from 2019 to 2024. Their median (P25, P75) total scores increased from 53.0 (49.8, 68.3) in 2019 to 86.5 (78.0, 96.0) in 2024, with normal morphology scores rising from 21.6 (16.4, 47.0) to 73.0 (54.0, 98.0). This demonstrates the value of training and sustained EQA engagement in reducing interlaboratory variability. In the future, differences caused by the subjectivity of different technicians can be reduced through instrument testing. Furthermore, whether manual or instrument detection is used, there is a need for accurate definitions of sperm morphology. However, neither the 5th7 and 6th17 editions of the WHO manual nor ISO 2316220 provides strict definitions for each specific defect. Therefore, future research should aim to precisely define normal sperm morphology and various types of abnormal sperm morphology. There is also a need for the clarification of which types of defects are related to infertility to provide a stronger basis for clinical diagnosis and treatment.
The method of using images to conduct EQA has long been applied in the field of blood cell morphology testing. According to the literature, the use of digital virtual slides has many advantages compared with traditional smears.21 However, in terms of the EQA of sperm morphology, there are relatively few reports on the use of images,22 while smears are currently more commonly used.23 It may not be possible to accurately determine whether the identification of each sperm defect type is correct when using smears, while images can make up for this deficiency and provide a beneficial supplement.
The present study has several strengths. A large number of sperm images (n = 450) were distributed. Moreover, these images almost covered all defect types described in the 6th edition of the WHO manual.17 These results enabled the determination of the correct rate for each sperm defect type. For defect types with a relatively low correct rate, more targeted training is needed. However, rare abnormalities (e.g., acephalic sperm, n = 1) lacked sufficient data for robust statistical analysis. Future EQA programs may expand the representation of rare sperm defects.
CONCLUSIONS
Our 5-year nationwide EQA study highlights substantial challenges in sperm morphology assessment, notably in the accurate recognition of normal sperm forms and specific defects of the sperm head. The consistently low correct rate for normal morphology (51.1%) indicates a tendency toward overly stringent criteria, which may contribute to the overdiagnosis of teratozoospermia. Moreover, the observed difficulties in classifying specific sperm abnormalities, particularly those affecting the head, underscore the urgent need for more precise standardized definitions and improved training protocols. These results suggest the need for a recalibration of sperm morphological assessment criteria. They also support integrating image-based EQA as a dual-purpose tool for evaluation and education, thereby enhancing diagnostic precision in male infertility. Collectively, these measures are essential in reinstating the clinical value of sperm morphology assessment in guiding infertility diagnosis and therapeutic strategies.
AUTHOR CONTRIBUTIONS
YLJ performed the statistical analysis and interpreted the data. YZ drafted the article and revised it critically for important intellectual content. FPL developed the original concept of this study. YBW, LY, YL, JTY, CCD, and YM processed the samples and collected data. All authors contributed to critical discussions, and read and approved the final manuscript.
COMPETING INTERESTS
All authors declare no competing interests.
ACKNOWLEDGMENTS
This work was supported by the Sichuan Science and Technology Program (No. 2024NSFSC0647) and the Health Commission of Sichuan Province Medical Science and Technology Program (No. 24SYJS01).
Supplementary Information is linked to the online version of the paper on the Asian Journal of Andrology website.
REFERENCES
- 1.Pisarska MD, Casson PR, Cisneros PL, Lamb DJ, Lipshultz LI, et al. Fertilization after standard in vitro fertilization versus intracytoplasmic sperm injection in subfertile males using sibling oocytes. Fertil Steril. 1999;71:627–32. doi: 10.1016/s0015-0282(98)00538-x. [DOI] [PubMed] [Google Scholar]
- 2.Van Waart J, Kruger TF, Lombard CJ, Ombelet W. Predictive value of normal sperm morphology in intrauterine insemination (IUI): a structured literature review. Hum Reprod Update. 2001;7:495–500. doi: 10.1093/humupd/7.5.495. [DOI] [PubMed] [Google Scholar]
- 3.Guzick DS, Overstreet JW, Factor-Litvak P, Brazil CK, Nakajima ST, et al. Sperm morphology, motility, and concentration in fertile and infertile men. N Engl J Med. 2001;345:1388–93. doi: 10.1056/NEJMoa003005. [DOI] [PubMed] [Google Scholar]
- 4.Jedrzejczak P, Taszarek-Hauke G, Hauke J, Pawelczyk L, Duleba AJ. Prediction of spontaneous conception based on semen parameters. Int J Androl. 2008;31:499–507. doi: 10.1111/j.1365-2605.2007.00799.x. [DOI] [PubMed] [Google Scholar]
- 5.Erdem M, Erdem A, Mutlu MF, Ozisik S, Yildiz S, et al. The impact of sperm morphology on the outcome of intrauterine insemination cycles with gonadotropins in unexplained and male subfertility. Eur J Obstet Gynecol Reprod Biol. 2016;197:120–4. doi: 10.1016/j.ejogrb.2015.12.014. [DOI] [PubMed] [Google Scholar]
- 6.Gatimel N, Moreau J, Parinaud J, Léandri RD. Sperm morphology: assessment, pathophysiology, clinical relevance, and state of the art in 2017. Andrology. 2017;5:845–62. doi: 10.1111/andr.12389. [DOI] [PubMed] [Google Scholar]
- 7.World Health Organization. WHO Laboratory Manual for the Examination and Processing of Human Semen. 5th ed. Geneva: World Health Organization; 2010. [Google Scholar]
- 8.Lemmens L, Kos S, Beijer C, Brinkman JW, van der Horst FA, et al. Predictive value of sperm morphology and progressively motile sperm count for pregnancy outcomes in intrauterine insemination. Fertil Steril. 2016;105:1462–8. doi: 10.1016/j.fertnstert.2016.02.012. [DOI] [PubMed] [Google Scholar]
- 9.van den Hoven L, Hendriks JC, Verbeet JG, Westphal JR, Wetzels AM. Status of sperm morphology assessment: an evaluation of methodology and clinical value. Fertil Steril. 2015;103:53–8. doi: 10.1016/j.fertnstert.2014.09.036. [DOI] [PubMed] [Google Scholar]
- 10.Kohn TP, Kohn JR, Ramasamy R. Effect of sperm morphology on pregnancy success via intrauterine insemination: a systematic review and meta-analysis. J Urol. 2018;199:812–22. doi: 10.1016/j.juro.2017.11.045. [DOI] [PubMed] [Google Scholar]
- 11.Danis RB, Samplaski MK. Sperm morphology: history, challenges, and impact on natural and assisted fertility. Curr Urol Rep. 2019;20:43. doi: 10.1007/s11934-019-0911-7. [DOI] [PubMed] [Google Scholar]
- 12.Villani MT, Morini D, Spaggiari G, Falbo AI, Melli B, et al. Are sperm parameters able to predict the success of assisted reproductive technology? A retrospective analysis of over 22,000 assisted reproductive technology cycles. Andrology. 2022;10:310–21. doi: 10.1111/andr.13123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Patel P, Carrasquillo R, Madhusoodanan V, Dadoun S, Patel A, et al. Impact of abnormal sperm morphology on live birth rates following intrauterine insemination. J Urol. 2019;202:801–5. doi: 10.1097/JU.0000000000000288. [DOI] [PubMed] [Google Scholar]
- 14.Kovac JR, Smith RP, Cajipe M, Lamb DJ, Lipshultz LI. Men with a complete absence of normal sperm morphology exhibit high rates of success without assisted reproduction. Asian J Androl. 2017;19:39–42. doi: 10.4103/1008-682X.189211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang Y, Yang J, Jia Y, Xiong C, Meng T, et al. Variability in the morphologic assessment of human sperm: use of the strict criteria recommended by the World Health Organization in 2010. Fertil Steril. 2014;101:945–9. doi: 10.1016/j.fertnstert.2013.12.047. [DOI] [PubMed] [Google Scholar]
- 16.Lemmens L, van den Hoven L, van Vrouwerff NJ, Braat DD, Nelen WL, et al. External quality control and training of semen analysis in the Netherlands: starting point for further reduction of outcome variability. Asian J Androl. 2022;24:15–20. doi: 10.4103/aja.aja_46_21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.World Health Organization. WHO Laboratory Manual for the Examination and Processing of Human Semen. 6th ed. Geneva: World Health Organization; 2021. [Google Scholar]
- 18.Jia YL, Wu YB, Yu L, Zheng Y, Yang TT, et al. Normal sperm head morphometric reference values in fertile Asian males. Asian J Androl. 2024;26:315–20. doi: 10.4103/aja202356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kohn TP, Kohn JR, Lamb DJ. Role of sperm morphology in deciding between various assisted reproduction technologies. Eur Urol Focus. 2018;4:311–3. doi: 10.1016/j.euf.2018.07.029. [DOI] [PubMed] [Google Scholar]
- 20.ISO (the International Organization for Standardization). ISO 23162:2021. Basic Semen Examination–Specification and Test Methods. 2021. [[Last accessed on 2023 Jul 31]]. Available from: https://www.iso.org/standard/74800.html .
- 21.Brereton ML, de La Salle B, Burthem J, Ardern J, Hickman L, et al. Review of the UK NEQAS (H) digital morphology pilot scheme for continuing professional development accessed via the internet. Int J Lab Hematol. 2008;30:365–71. doi: 10.1111/j.1751-553X.2008.01086.x. [DOI] [PubMed] [Google Scholar]
- 22.van den Hoven L, van Vrouwerff NJ, Dijkstra IM, Weeteling J, Brinkman JW, et al. Evaluation of structural problems in the application of strict criteria for sperm morphology assessment. Andrology. 2025;13:512–8. doi: 10.1111/andr.13684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zheng Y, Wu Y, Yu L, Jia Y, Yang T, et al. How has the external quality assessment/proficiency testing of semen analysis been developed in the past 34 years: a review. Clin Chem Lab Med. 2024;63:670–83. doi: 10.1515/cclm-2024-1062. [DOI] [PubMed] [Google Scholar]