How to count sperm properly : checklist for acceptability of studies based on human semen analysis

Transcription

1 Human Reproduction, Vol.31, No.2 pp , 2016 dvanced ccess publication on December 18, 2015 doi: /humrep/dev305 OPINION How to count sperm properly : checklist for acceptability of studies based on human semen analysis Lars Björndahl 1, *, Christopher L.R. Barratt 2, David Mortimer 3, and Pierre Jouannet 4 1 Centre for ndrology and Sexual Medicine, Karolinska University Hospital and Karolinska Institutet, Huddinge C2:94, S Stockholm, Sweden 2 Reproductive Medicine, University of Dundee, Dundee, UK 3 Oozoa Biomedical Inc., Vancouver, BC, Canada 4 Université Paris Descartes, Paris, France *Correspondence address. Lars.Bjorndahl@ki.se Submitted on October 19, 2015; resubmitted on October 19, 2015; accepted on November 17, 2015 study question: Can a tool be developed for authors, reviewers and editors of the ESHRE Journals to improve the quality of published studies which rely on semen analysis data? summary answer: basic checklist for authors, reviewers and editors has been developed and is presented. what is known already: Laboratory work which includes semen analysis is burdened by a lack of standardization. This has significant negative effects on the quality of scientific and epidemiological studies, potential misclassification of patients and the potential to impair clinical treatments/diagnoses that rely on accurate semen quality information. Robust methods are available to reduce laboratory error in semen analysis, inducing adherence to World Health Organization techniques, participation in an external quality control scheme and appropriate training of laboratory personnel. However, journals have not had appropriate systems to assess if these methods have been used. study design, size, duration: fter discussion at a series of ssociate Editor Meetings of the ESHRE Journals the authors of the present text were asked to propose a tool for authors, reviewers and editors of the ESHRE Journals to ensure a high quality assessment of submitted manuscripts which rely on semen analysis data, including a detailed verification of the relevance and the quality of the methods used. participants/materials, setting, methods: N/. main results and the role of chance: basic checklist for authors, reviewers and editors is presented. The checklist contains key points which should be considered by authors when designing studies and which provides essential information for when the submitted manuscript is evaluated. For published articles the answers in the checklist are suitable to be available as supplementary data, which will also reduce the space necessary for technical details in the printed article. limitations, reasons for caution: Guidelines such as these should not be used uncritically. It is therefore important that submitting authors, in situations where their study does not comply with the basic requirements for semen analysis, not only explain all methodological deviations but also declare the level of uncertainty in their analyses and how it complies with, or might confound, the aims of the study. wider implications of the findings: The fundamental importance of appropriate and robust methodology to facilitate advances in scientific understanding and patient management and treatment, is now accepted as being paramount. Use of the semen analysis checklist should be part of this process, and when completed and signed by the corresponding author at the time of submitting a manuscript should result in greater transparency, and ultimately uniformity. It is hoped that this initiative will pave the way for wider adoption of the methodology/reporting by other biomedical, epidemiological and scientific journals, and ultimately become the standard of practice for papers reporting semen analysis results obtained in laboratory and clinical andrology. Systems to assist referees, authors and editors to present high quality findings should have a significant impact on the field of reproductive medicine. study funding/competing interest(s): No funding was obtained for this work. The authors have no competing interests in relation to the present publication and checklist. trial registration number: N/. Key words: semen analysis / standardization / laboratory quality / evidence-based medicine / check-list / study design / manuscript review & The uthor Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. ll rights reserved. For Permissions, please journals.permissions@oup.com

2 228 Björndahl et al. Introduction Laboratory andrology, which includes semen analysis, is a branch of reproductive science burdened by severe lack of standardization (Harvey and Jackson, 1945; Tomlinson, 2010; Walczak-Jedrzejowska et al., 2013) that has caused it to lag behind in the development of sophisticated testing methods and has also led to suboptimal care for childless couples. It is therefore immensely important for scientific journals publishing studies in the field of human reproduction to promote methodological standardization that will lead to more accurate and robust information, leading directly to improved quality in laboratory andrology, and hence in scientific studies and reproductive services. This is even more urgent when considered from the perspective of evidence-based medicine: with poor laboratory methodologies, not even the best-designed clinical study can provide reliable and valuable results. The simple explanation for this is that RCTs and other clinical studies designed to produce evidence are rarely focused on laboratory services; their aim being primarily to evaluate the most efficacious treatment options, not the quality or reliability of the laboratories providing basic diagnostic testing. However, if laboratories produce results that often suffer from substantial, more or less random, errors this will lead to the more or less random diagnostic classification of patients, andclearlyeven well designedbasic or clinical studies will fail to produce useful information. The Special Interest Group in ndrology (SIG) was formed in 1992duringtheEuropeanSocietyofHumanReproductionandEmbryology (ESHRE) nnual Meeting in the Hague under the leadership of Professor Lynn Fraser. In addition to the usual activities for an ESHRE special interest group (organizing Campus Courses and Pre- Congress Courses), the SIG presented in 1994 a course on Basic Semen nalysis ( BS Course ) that has since been standardized and repeated in a many countries (Sweden, The Netherlands, Belgium, Denmark, Norway, Finland, Ukraine, Spain, Greece, Portugal,Poland,Italy,Southfrica,Canada,andtheUK)inarangeoflanguages (Björndahl et al., 2002). The BS course has been revised to comply with relevant recommendations from the World Health Organization (WHO) (World Health Organization, 2010) andsig policy decisions (Björndahl et al., 2010; Barratt et al., 2011). In addition to the BS Course, the SIG has operated an external quality assurance (EQ) scheme since 1999, and a comprehensive guideline for practically all aspects of designing, accomplishing and publishing scientific studies involving assessments of semen quality SEMQU was published recently (Sanchez-Pozo et al., 2013). The comprehensive SEMQU document clarifies the necessary steps and considerations for investigations aiming at robust and reliable determination of semen parameters. There are, however, studies using semen analysis results for other purposes, for example the evaluation of treatment modalities or prognostic usefulness of new techniques and/or different methods. During the 2012 ssociate Editor Meetings of the ESHRE Journals in Istanbul the authors of the present text were asked to propose a tool for authors and reviewers of the ESHRE Journals to ensure a high quality assessment of submitted manuscripts which rely on basic semen analysis data, including a detailed verification of the relevance and the quality of the methods used. There are several important reasons for the ESHRE Journals to strengthen their requirements for data on semen analysis, as described below. General considerations of the quality of science Laboratory investigations performed with a high level of uncertainty i.e. subject to a large influence of random errors can conceal true biological relationships. t the simplest level, a lack of significant correlation with an investigated factor might be due to a high level of uncontrolled errors that led to inaccurate, and hence highly variable, observations. The fundamental importance of proper and robust methodology for sound scientific development in medicine and biomedical science is gaining substantial interest and being widely discussed (Plant et al., 2014). Much attention has already been focused on the appropriate selection and size of experimental and control groups (in terms of achieving necessary statistical power) but all efforts to obtain these essential requirements are foredoomed if the laboratory investigations lack sufficient robustness, reliability and repeatability. Implementation of new methods in other centres Laboratory methods must be described properly to permit other centres to implement the techniques and/or repeat a study to verify its conclusions. This also includes the proper participation by both the publishing and implementing centres in EQ programmes that test the critical parts of semen analysis involved. Consistency in reporting embryology and andrology laboratory results Consistent requirements for the stringent reporting of embryology and andrology laboratory data will hopefully emphasize the necessity of implementing adequate laboratory techniques in all centres. Furthermore, the awareness of, and compliance with, modern laboratory andrology standards and techniques is often lacking. For instance, it appears that many authors are unaware that the WHO main recommendation for the assessment of human sperm morphology is actually based on the Tygerberg Strict Criteria (World Health Organization, 2010), quite often resulting in incorrect comments in manuscripts on what are supposed to be the same assessment criteria but described and presented as separate methods. Furthermore, the present WHO recommendations do not include motility and concentration assessments carried out on raw semen by computer-assisted sperm analysis (CS) (ESHRE Special Interest Group in ndrology, 1998; World Health Organization, 2010), although statements like semen analysis was performed according to WHO using CS occur frequently in submitted manuscripts despite their being clear non sequiturs. The purpose of this paper is to provide guidelines for all scientists designing, performing and writing reports on investigations that rely on semen analysis data that are to be submitted to any of the ESHRE Journals. The concept is that the checklist in the ppendix should be compulsory, and must be completed and signed by the corresponding author at the time of submitting the manuscript. ny deviation from this guideline should be declared and explained in the Materials and Methods section of the manuscript (see ppendix). The checklist would also help reviewers and Editors in evaluating the submitted manuscripts.

3 Obtaining accurate data from human semen analyses 229 International standards for medical laboratory testing In all areas of medical laboratory testing the principles of ISO (International Standards Organization, 2012) define the basic requirements for quality as well as competence. Consequently, any semen analysis that is performed for diagnostic reasons, or during a fertility treatment (e.g. IVF in an embryology lab), must be expected to conform to these international standards. Section 5.1 of the ISO elaborates on Personnel expectations that adequate training shall have been provided, and that competence shall also have been assessed and established according to established criteria, and also will be re-assessed on a regular basis. Section 5.4 covers Pre-examination Processes, including sample collection and handling (along with the pertinent pre-collection and collection instructions to patients), specimen transportation, specimen reception and handling, preparation and storage. Section 5.5 covers Examination Procedures, including their validation and the need to establish the measurement uncertainty of results. Section 5.6 considers how to ensure the quality of results, including requirements concerning internal quality control (IQC) and EQ. Of particular importance with regard to EQ activities, which can include the analysis of materials such as verified reference materials, exchange of samples with other laboratories, and control materials from interlaboratory comparison programmes, are the requirements of ISO/IEC 17043:2010 (International Standards Organization, 2010). typical feature of EQ programmes is to provide education to participants and promote quality improvement: hence advisory and educational comments should comprise part of the report returned to participants to achieve this aim. For semen analysis this means that the distributed EQ specimens must have assigned values that allow assessment of accuracy (trueness and precision) as well as the measurement uncertainty required or expected for each measurand. When a consensus value is used as the assigned value (see nnex B of ISO/IEC 17043:2010), the proficiency testing provider (EQ scheme operator) must document the reason for that selection and also estimate the uncertainty of the assigned value in order to establish a reasonable target range (Palacios et al., 2012). The ESHRE SIG EQ scheme assigns such reference values based on the average of results and variation reported by its reference laboratories those that have been recognized as operating to high standards in accordance with the recommended methodology (Björndahl et al., 2010) and staff training and proficiency. Consequently, if a participating laboratory discovers that its values deviate significantly from the reference values, then corrective action can be implemented using the EQ scheme reference values as the target. However, a scheme that only reports participating laboratories performance in comparison to each other, with no reference to any reference values (e.g. using a statistic such as the all laboratories trimmed mean ), embodies no capacity for real world quality improvement in terms of trueness. The system of operation of the EQ scheme is therefore essential for quality of the results obtained in the individual participating laboratories. Thus, participation in schemes running with suboptimal methodology will require more detailed considerations of uncertainty (correctness and variability) if the centre wishes to publish data. IQC activities among laboratory staff need to be anchored by a reference staff member, whose training and competency have been verified against a properly validated EQ scheme that incorporates effective quality improvement capabilities, as described above. Until all laboratories performing semen analysis conform to these standard international expectations for medical testing laboratories, semen analysis data reported from studies on human semen quality will remain of unknown quality and hence continue to be suspect. daptation to investigations of different types The checklist provided in the ppendix to this paper is based on the WHO and ESHRE-SIG recommendations for basic semen analysis, and conforms to the standards taught in the ESHRE-SIG BS Course (Barratt et al., 2011). While the level of uncertainty (error) that is acceptable can certainly vary, depending on the purpose of the investigation, it should not as a general rule exceed +10%. But guidelines such as these should not be used uncritically. For example, in large epidemiological population studies a high number of patients or samples can compensate for the influence of random errors in each individual assessment when calculating average values for populations, in contrast to the case where a single result from one individual is obtained with the intent of determining further investigations or selecting the best treatment modality. However, in an experimental study where different aliquots of each semen sample are treated in different ways and compared, poor laboratory methods can conceal true differences: paired analyses will be subject to comparable uncertainty of measurement errors. Here each sample acts as its own control and pair-wise or similar statistics with higher power should be used, sometimes allowing the observation or (discovery) of even quite small effects in spite of suboptimal laboratory investigations. On the other hand, in the clinical situation where laboratory data from the investigation of one or a few semen analyses will be used to decide a treatment modality, there may be a need for even higher reliability to avoid the wrong treatment being given to the couple. n everyday example to illustrate this would be where a clinical laboratory makes decisions on whether to perform routine IVF or ICSI based on a sperm concentration obtained after sperm preparation by counting in a shallow chamber (e.g. a Makler chamber) using a standard procedure leading to too few observations in order to obtain a reliable result. Such simplified procedures are still common in clinical practice although they are not in compliance with published recommendations or even those of the manufacturer of the device (Makler, 1978, 1980). For instance, if ICSI might be deemed necessary when the sperm concentration after preparation is,0.8 million/ml (M/ml), ordinary IVF treatment when the concentration is.1 M/ml, and in cases where the concentration is believed to be between 0.8 and 1 M/ml, half of the oocytes are injected (ICSI) and the other half given ordinary IVF treatment (an IVF/ICSI split ). This may sound reassuring, but how reliable is the basis for this decision? In such counts, the uncertainty of the result depends on how many spermatozoa were observed. Consider a case where the assessment gives the result 0.9 (M/ml) (Fig. 1). This would be a clear case for an IVF/ICSI split. However, taking into consideration the huge influence of random errors when only a few

4 230 Björndahl et al. Figure 1 schematic representation of the uncertainties associated with different numbers of spermatozoa counted in shallow or deep counting chambers, based on examples of different methods actually used in some IVF centres to choose between different treatment strategies: IVF only, ICSI only or with results of sperm counting in the borderline zone (grey box) 50:50 IVF/ICSI. The proportion of each area-under-the-curve that falls within the grey box represents the proportion of results that will lead to the intended treatment modality. The curves show the probable distribution (Poisson) of the true sperm concentration in a sperm preparation for assisted reproductive techniques, based on the number of counted spermatozoa. For the shallow (e.g. Makler) chamber (red curve), the 9 spermatozoa were counted in 10 squares, a practice quite common among many IVF centres, although not according to published recommendations (Makler, 1978, 1980). When 100 spermatozoa are counted in a haemocytometer (yellow curve) the uncertainty decreases as illustrated by the narrower peak. ssessment of 400 spermatozoa (green curve) would mean that 95% of counts giving the result 0.9 millions/ml (M/ml) would have the true value within the interval M/ml. In the case with 9 counted spermatozoa only, a minute fraction of the true values will be within the expected interval (grey box), resulting in a more or less random selection of treatment. majority of couples would, in this example, actually receive a treatment that was not intended. spermatozoa are counted, only one in four patients will be offered the treatment that is believed to be appropriate based on the assessment and the set limits. In over 40% of the couples with the measurement 0.9 M/ml, the true value is.1 M/ml and these couples should have been given IVF only. In over 30% of these couples the true sperm concentration would be,0.8 M/ml for which the intended treatment was ICSI only. clinic using WHO and ESHRE standards for semen analysis would give 97% of all patients the intended treatment and the influence of random error would be kept under 5% (Fig. 1), calculations that are not always obvious to those lacking specialist insights into quantitative laboratory science. Clearly this is an area for improvements concerning patient securityand efficiency in the treatment of infertility; comprehensive education in laboratory andrology should therefore be an essential component in the development of reproductive medicine. Furthermore, data obtained using suboptimal techniques is not warranted in scientific studies. It is therefore important that submitting authors, in situations where their study does not comply with the basic requirements for semen analysis, not only explain all methodological deviations but also declare the level of uncertainty in their analyses and how it complies with or might confound the aims of the study. When recommendations on clinical decision limits are given in relation to an investigation, information should be provided as to not only the positive and negative predictive values but also the uncertainty of recommended analytical methods around the suggested clinical decision limit. It is not sufficient to present a simple cut-off obtained from a receiver operating characteristic curve (ROC curve) analysis. Conclusions To support scientific development in Reproductive Medicine, as well as improve patient diagnosis, management and treatment, it is essential that standards be raised in regard to the publishing of results from studies involving semen analyses. We therefore propose that the basic checklist shown in the ppendix of this paper be used by submitting authors as well as peer reviewers and Editors of all pertinent journals. Based on discussions with Editors and ssociate Editors of the three ESHRE journals there are several strategies for actually implementing this within the submission and review process. Clearly the Questionnaire would need to be completed and signed byat least the submitting authorat the time of initial manuscript (MS) submission. It should probably also be made clear that making a false declaration would constitute scientific fraud. The process for review of a submitted MS could then be carried out in different ways, depending on the individual circumstances of the MS. In some cases, a technical review of the MS could be done by an Expert Panel appointed by the journal to verify conformity (akin to statistical review such as is undertaken for meta-analyses) before sending the MS to reviewers, while in other cases an Expert Panel view may be pertinent before the ssociate Editor s decision. Simple expediency would suggest doing this step post-triage, to reduce the load on the Expert Panel. To do a technical review before sending an MS to peer reviewers (i) would allow authors to resolve perceived non-conformities prior to review, and (ii) facilitate the reviewers work. For the future, we hope that this initiative will be adopted by other biomedical, epidemiological and scientific journals, and ultimately become

5 Obtaining accurate data from human semen analyses 231 the standard of practice for papers reporting results obtained in laboratory and clinical andrology. uthors roles ll authors have contributed equally to the conception and writing of the manuscript. Funding No funding was obtained for this work. Conflict of interest None declared. References Barratt CL, Bjorndahl L, Menkveld R, Mortimer D. ESHRE special interest group for andrology basic semen analysis course: a continued focus on accuracy, quality, efficiency and clinical relevance. Hum Reprod 2011;26: Björndahl L, Barratt CL, Fraser LR, Kvist U, Mortimer D. ESHRE basic semen analysis courses : immediate beneficial effects of standardized training. Hum Reprod 2002;17: Björndahl L, Mortimer D, Barratt CLR, Castilla J, Menkveld R, Kvist U, lvarez JG, Haugen TB. Practical Guide to Basic Laboratory ndrology, 1st edn. Cambridge: Cambridge University Press, ESHRE Special Interest Group in ndrology. Guidelines on the application of CS technology in the analysis of spermatozoa. In: Mortimer S, Mortimer D, Fraser L (eds). Hum Reprod, 1998, Harvey C, Jackson MH. ssessment of male fertility by semen analysis an attempt to standardise methods. Lancet 1945;2: International Standards Organization. ISO/IEC 17043:2010 Conformity ssessment General Requirements for Proficiency Testing, Geneva, International Standards Organization. ISO 15189:2012 Medical Laboratories Requirements for Quality and Competence, Geneva, Makler. new chamber for rapid sperm count and motility estimation. Fertil Steril 1978;30: Makler. The improved ten-micrometer chamber for rapid sperm count and motility evaluation. Fertil Steril 1980;33: Mortimer D, Barratt CLR, Björndahl L, de Jager C, Jequier M, Muller CH. What should it take to describe a substance or product as sperm-safe. Hum Reprod Update 2013;19:i1 i45. Palacios ER, Clavero, Gonzalvo MC, Rosales, Mozas J, Martinez L, Ramirez JP, Björndahl L, Morancho-Zaragoza J, Fernandez-Pardo E et al. cceptable variability in external quality assessment programmes for basic semen analysis. Hum Reprod 2012;27: Plant L, Locascio LE, May WE, Gallagher PD. Improved reproducibility by assuring confidence in measurements in biomedical research. Nat Methods 2014;11: Sanchez-Pozo MC, Mendiola J, Serrano M, Mozas J, Bjorndahl L, Menkveld R, Lewis SE, Mortimer D, Jorgensen N, Barratt CL et al. Proposal of guidelines for the appraisal of SEMen QUlity studies (SEMQU). Hum Reprod 2013; 28: Tomlinson M. Is your andrology service up to scratch? Hum Fertil (Camb) 2010;13: Walczak-Jedrzejowska R, Marchlewska K, Oszukowska E, Filipiak E, Bergier L, Slowikowska-Hilczer J. Semen analysis standardization: is there any problem in Polish laboratories? sian J ndrol 2013;15: World Health Organization. WHO laboratory manual for the examination of human semen and sperm-cervical mucus interactions, 4th edn. Cambridge, UK: Cambridge University Press, World Health Organization. WHO laboratory manual for the examination and processing of human semen, 5th edn. Geneva: World Health Organization, ppendix check list for authors, reviewers and editors ny deviation from this guideline (i.e. a box without a tick) must be declared and explained in the Materials and Methods section of the manuscript, including explaining its effect on the measurement uncertainty of the data, in order toallow the reader toevaluatethe qualityofthe analysesperformed. Investigations that would be subject to this requirement can roughly be classified as clinical (evaluating patient treatment, diagnostic classification or predictive powers of certain assessments), experimental (e.g. exposure of sperm to different compounds or in vitro treatments (Mortimer et al., 2013)), or epidemiological (evaluating variations in semen characteristics or effects of exposure populations to certain compounds or other circumstances). ny scientific rationale for not complying with the guidelines, which is not included in the Materials and Methods section of the manuscript, must be substantiated to Editor and reviewers. Patients For clinical studies: The patient population (e.g. patients, volunteers, students) has been declared in the manuscript, together with the recruitment method and inclusion and exclusion criteria. If the study concerns couples being investigated for infertility then the following must be specified in the manuscript: fertility status of female partner; and primary, secondary or other level of investigation of the man. If used in the manuscript, the term male factor must be completely defined. General aspects Patients were instructed to maintain 2 7 days of sexual abstinence before collecting a sample for investigation. Patients were informed about the importance of reporting any missed early ejaculate fractions, and men s answers were noted on the laboratory form. For specimens not collected at the laboratory, patients were instructed to avoid cooling or heating of the semen sample during transport to the laboratory. Samples were kept at 378C before initiation of and during the analysis in case of sperm motility assessment. For samples collected adjacent to the laboratory, analysis was initiated after completion of liquefaction and within 30 min after ejaculation. If this was not done and more importantly when some of the samples are collected in the laboratory and others are collected at home it should be checked that this did not influence the data (and, if yes, that this effect must be included as a confounding factor in the statistical analysis). Liquefaction was first checked within 30 min after ejaculation. Volume was determined either by weighing or using a wide-bore volumetric pipette. Viscosity was measured using either a wide-bore pipette or a glass rod.

6 232 Björndahl et al. ll staff members who performed the analyses have been trained in basic semen analysis (ESHRE Basic Semen nalysis Course or equivalent and further in-house training) and participate regularly in internal quality control. If more than one method can be recommended for a particular characteristic (e.g. to measure volume), only one should be used in a given study. Sperm concentration assessment Semen aliquot to be diluted for sperm concentration assessment was taken with a positive displacement pipette (i.e. a PCR pipette ) using a recommended diluent (state which diluent: ). Only standard dilutions were used (1:50, 1:20, or 1:10). Sperm concentration was assessed using haemocytometers with improved Neubauer ruling. Haemocytometers were allowed to rest for min in a humid chamber to allow sedimentation of the suspended spermatozoa onto the counting grid before counting. Sperm counting was done using phase contrast microscope optics ( ). Comparisons were made between duplicate counts, and counts re-done when the difference exceeded the acceptance limits. Typically at least 200 spermatozoa were counted in each of the duplicate assessments. Sperm motility assessment Motility assessments were performed at 378C C. Motility assessments were done using phase contrast microscope optics ( ). Sperm motility was classified using a four-category scheme: rapid progressive, slow progressive, non-progressive, and immotile (World Health Organization, 1999; Björndahl et al., 2010; Barratt et al., 2011). Motility assessments were done in duplicate and compared; counts were re-done on new preparations when the difference between duplicates exceeded the acceptance limits. The wet preparation was made with a drop of ml and a x mm coverslip to give a depth of mm (must be at least 10 mm, but not too deep so as to allow spermatozoa to move freely in and out of focus; typically ca. 20 mm). t least 200 spermatozoa were assessed in each duplicate motility count. t least 5 microscope fields of view were examined in each duplicate count. Sperm vitality assessment validated supravital staining, appropriate to the type of microscope optics utilized, was used to assess sperm vitality. t least 200 spermatozoa were evaluated in each sample. ssessments were done under high magnification ( ) using a 100 high resolution oil immersion objective and bright field microscope optics (Köhler illumination). Sperm morphology assessment Tygerberg Strict Criteria were used for the evaluation of human sperm morphology. Note: nother classification could be used for scientific studies with specific aims if the classification is described or referenced. Depending on the aim of the study, the evaluation of particular abnormal forms might be useful. bnormalities are recorded for all four regions of the spermatozoon (head, neck/midpiece, tail and cytoplasmic residue) and the Teratozoospermia Index or TZI was calculated (Björndahl et al., 2010; Barratt et al., 2011). If the laboratory claims to use Tygerberg Strict Criteria for the evaluation of human sperm morphology, then the laboratory must participate in an external quality assurance scheme to verify that its assessments comply with these criteria. The Papanicolaou staining method adapted for the assessment of human sperm morphology was used. For specific aims other staining methods could be used, but must then be declared and explained. t least 200 spermatozoa were assessed in each ejaculate. ssessments were done under high magnification ( ) using a 100 high resolution oil immersion objective and bright field microscope optics (Köhler illumination). Other findings The presence of abnormal clumping (aggregates and agglutinates) was recorded. bnormal viscosity was recorded. The presence of inflammatory cells was recorded and reported if more than 1 million/ml. For the purpose of classifying infertility status (World Health Organization, 2010), antisperm antibodies were examined with a validated screening test (state which method was used: ). nalysing data The actual duration of sexual abstinence (in hours or days ) was recorded for each sample and included in the data reported in the manuscript. s a minimum in clinical studies, semen volume, sperm concentration, total number of spermatozoa/ejaculate, and abstinence time are given to reflect sperm production and output; only samples identified as having been collected completely can be included in the study. Confounding factors have been considered for statistical analysis: e.g. abstinence time and age, to evidence secular or geographical variations in sperm concentration or sperm count. If appropriate, optional biochemical markers for prostatic, seminal vesicular and epididymal secretions were analysed and reported both as concentration and total amount. Signs of active infection/inflammation were noted and considered in the analysis of data in the study (e.g. inflammatory cells, impaired sperm motility, possibly also antisperm antibodies and reduction of secretory contributions).