{"paper_id":"753c710b-d49a-433b-bd8f-002398f6aec7","body_text":"Theoretically, our article draws on the STS insight that knowledge both\ndescribes and constitutes its objects (e.g.  Jasanoff, 2004 ;  Knorr-Cetina,\n1981 ;  Latour and Woolgar, 1986 ). Accordingly, we conceptualize\nembryo culture as ways of knowing about embryonic life and viability. We aim\nto explore technological processes and practices of knowledge production and\nhow they produce (what is perceived to be) legitimate and plausible\nknowledge about embryos and their viability.\nKnowledge production in the laboratory is generally seen as ‘through and\nthrough relational, embodied, affective and practical’ craftwork ( Meskus, 2018 :\n108). It is work that involves a multitude of actors, local and global,\nhuman and non-human. In IVF laboratories those actors include technologies,\nstandards, embryologists, other clinical staff, intended parents, national\nand supranational laws and regulations on the medical use of human tissues\nand cells, the industrial players and the living biomaterial itself (see\n Helosvuori,\n2019 ). It is hands-on work, but ‘there can be no separation of\nthe hands-on and the intellectual reshaping of what cells are [and] what\nthey can do’ ( Landecker, 2007 : 26), as shown by pioneering laboratory\nstudies (e.g.  Knorr-Cetina, 1981 ;  Latour and Woolgar, 1986 ;  Lynch, 1985 ). As\n Sennett\n(2008)  describes, craft is characterized by affective and\nattentive engagement with its object material to fulfill ‘the desire to do a\njob well for its own sake’.\nHowever, craftwork is subject to increasing political and economic demands for\nits transformation into more standardized and automated systems of knowledge\nproduction, moving from small-scale craft-like surroundings to\nindustrialized settings – scaling up. While standards and protocols may be\nmore common in clinical work than in experimental research, they also become\nappropriated for uses for which they were never imagined in their design\n( Bowker and Star,\n1999 ;  de\nLaet and Mol, 2000 ), used in reflexive rather than clear-cut\nways, worked around ( Timmermans, 2015 ), and tinkered with to make them\n work  ( de Laet and Mol, 2000 ;  Timmermans and Epstein,\n2010 ).\nIn this article, we introduce a response to standardization that does not rest\non hype, hope or excitement. Some clinics resist standardization by\nreinstating aspects of the craft production of knowledge, in the process\nsketching new political realities. This response has previously been\nidentified in work practices such as artisanal food production, the home\nbirth movement ( Rothman, 2016 ) and even research management ( Davies and Horst,\n2015 ), but it has not been identified in laboratory work to\ndate.\nSTS approaches have historically been accused of being apolitical or siding\nwith the strong due to their inability to fully theorize how and to what\nends materials link to the capitalist mode of production or institutional\npower relations (e.g.  Braunmühl, 2018 ;  Law, 2004 : 13–14;  Mol and Mesman,\n1996 ;  Rekret, 2018 ). It can be argued, however, that STS approaches\nto knowledge production and its methods provide the tools to attend to\n‘goodness’ in various activities: caring for many participants\nsimultaneously including oneself (see  Davies and Horst, 2015 ;  Homanen, 2019 ;\n Lemke,\n2018 : 45;  Mol, 2008 ) and encompassing all participants’ economic\nconditions. This caring exists alongside the logics and interests of\neconomic value accumulation and instrumentalization.\n\nTo incorporate various laboratory knowledge production practices across\ndifferent locations, we use multi-sited ethnography ( Falzon, 2009 ;  Hine, 2007 ).\nMulti-sited ethnography recognizes that the research’s matter of concern is\nconstantly crafted in a range of locations and practices, including\nlaboratories, doctors’ offices, intended parents’ private reproductive\nvoyages and peer support groups. An ethnographic time frame is needed to\ndeal with the entire trajectories of embryo culture.\nThe article is based on two multi-sited ethnographic studies conducted at a\ntotal of eight different private fertility clinics in Finland during\n2013–2017, with a follow-up study from autumn 2019 onward. Both of our\nlarger research projects are concerned with the constitution of social\nrelations and selves in clinical reproductive healthcare practices in the\ncontext of healthcare marketization. The fieldwork sites encompass fertility\nclinic spaces including IVF laboratories, conferences and events and peer\nsupport groups for intended parents. During our fieldwork periods, there\nwere between ten and fourteen commercial fertility clinics in Finland in\ntotal. The combination of our data not only allows us to analyze material\nfrom eight of those commercial clinics, but also to observe the links,\nvariations and tensions among practices that would not have been visible in\nour respective separate data sets.\nAs is common in ethnographic enquiry, we repeatedly reframed our analysis of\nour material through new knowledge produced collaboratively with\nparticipants in the field ( Hammersley and Atkinson, 1995 ;\n Holmes and\nMarcus, 2008 ). This enabled us to examine how knowledge\npractices are realized and challenged in specific, situational, everyday\nclinical work. It also enabled us to see the political work involved in\nnormalizing technologies and in contesting, reformulating and reinforcing\ntechnological practices and the market.\nAltogether, the data collected in the two larger studies during 2013–2017 and\nin the follow-up study that started in autumn 2019, comprises ethnographic\nobservations and video recordings (105 videos of appointments and\nprocedures) from six private fertility clinics (clinics A, B, C, D, E, and\nF) and 27 in-depth interviews with medical personnel in seven clinics (eight\nclinics participated in total). The videotapes analyzed in-depth in this\narticle comprise recordings of 42 procedures with embryologists in the\noperating theater and five videos where embryologists are showing and\nexplaining the development of patients’ embryos to them via TLS. The\ninterviews analyzed comprise nine interviews with embryologists and six with\ndoctors, as they are the ones in charge of laboratory work in different\nways. The fieldwork was conducted in periods lasting between a few days and\ntwo weeks alongside an intense period of participant observation at one\nlaboratory for one month. Additionally, the fieldwork included participant\nobservation at five Finnish and international conferences for\nmedical/laboratory staff, five public events that brought together various\nstakeholders on (in)fertility issues and two information/recruitment events\nfor potential egg donors organized by private clinics. We also collected\ndocumentary material, such as legislative and regulatory documents, care and\nlaboratory protocols and guidelines and hand-outs distributed to intended\nparents.\nAlthough this article focuses on observations and interviews conducted with\nspecific medical staff in clinics, we also draw on data gathered from\nobservations conducted over seven months in four peer support groups for\npeople experiencing involuntary childlessness, where necessary. A Finnish\ninfertility association organized these groups. Thirteen interviews with\nintended parents were also conducted.\nAll interviewees and participants in video recordings signed an informed\nconsent form. Other participants gave oral consent to participate or were\ninformed beforehand of the presence of the researcher and given an\nopportunity to object or withdraw from the event. At public events and\nconferences, the researchers informed the organizers of the study\nbeforehand.\n\nThe legislation regulating infertility treatment in Finland dates to 2006, when\nthe Act on Assisted Fertility Treatments (1237/2006) was passed. Previously,\ntreatments had been self-regulated by professionals and their associations.\nThe law legalized gamete donation and fertility treatment for single women\nand (in effect) lesbian couples, but it criminalized surrogacy and banned\nanonymous donation and any remuneration of gamete donors. A state donor\nidentity register with an identity release system was established. All\ngamete and embryo donors since 2007 have been registered, and children born\nas a result of donor IVF may on request receive identity information about\nthe donor after the age of eighteen. Finland has been described as rather\npermissive ( Eriksson,\n2017 ). Indeed, it is so compared with some other European\ncountries, such as Norway and Germany, where egg donation is forbidden.\nIVF laboratories are further regulated by the Act on the Medical Use of Human\nOrgans, Tissues and Cells (101/2001), which has been extensively amended\nsince 2007 to align it with common European Union (EU) regulations regarding\nthe procurement, storage, transport and traceability of tissues and cells,\nand their screening for infection. These regulations were introduced by the\nEU in the Tissue and Cell Directives of 2004 and 2006 (2004/23/EC;\n2006/17/EC; 2006/86/EC). In effect, the EU legislation harmonizes procedures\nin IVF laboratories across Europe. However, there is no EU legislation on\nassisted reproduction in general and the overall picture in Europe has been\ndescribed as a patchwork ( ESHRE, 2017 ).\nAccording to the Tissue Act (101/2001) amended in 2007, procurement, testing,\nprocessing, preservation, storage and distribution must take place in a\nlicensed tissue establishment that is registered and regularly inspected by\nFinnish Medicines Agency (Fimea), a government agency. Additionally, tissue\nestablishments provide Fimea with annual reports. In general (and in line\nwith EU legislation), the Finnish legal requirements include a push toward\nstandardized auditing and quality systems in order to minimize risks to\nhuman health (see  European Commission, 2016 ). For traceability purposes,\nestablishments such as IVF laboratories must keep registers of all\npharmaceutical products and tissues used, the donors/patients to whom the\ntissue belongs and all procedures conducted. Furthermore, a register of\nadverse effects and hazardous situations needs to be kept. Many common\ntechnical requirements for the coding, processing, preservation, storage,\nand distribution of human tissues and cells are also documented in the EU\ndirectives and correspondingly referred to in Finnish law.\nFinland combines a state-funded Nordic welfare system with a growing commercial\ncare sector. The policy discourse of the centrality of economic\ncompetitiveness has been particularly strong in Finland since the 1990s\n( Mulinari et al.,\n2009 ): Public-private partnerships play a vital role in service\nprovision. The Finnish government even works in conjunction with the private\nsector to attract medical tourists; FinlandCare, a mediator organization\ncoordinating the sale of care services abroad, was established as a\ncollaboration between government agencies and private companies. Compared\nwith Sweden, for example, Finland now has a wide network of private\nhospitals and clinics. Finland has also become a destination for fertility\ntravel for donor eggs, mainly from other Nordic countries where laws,\npolicies or service systems are more restrictive (for a more detailed\ndescription of the transnational Finnish egg donation context, see  Homanen,\n2018 ).\nDuring our fieldwork, Finland has had 17–25 clinics with licenses to perform\nIVF treatments. Of these, 10–14 have been commercial establishments. Over\ntime, private enterprise has moved from maverick small businesses to a few\nlarge chains of clinics with links to international investors and branches\nin Russia and some Baltic countries. The consolidation of the sector in\nFinland can be seen in ongoing mergers and acquisitions: In a little over\n20 years, all but a couple of clinics have been merged with or acquired by\nanother clinic, a chain of clinics or a large healthcare company.\nWhile clinics in Finland thus seem to follow recent global trends in the\nindustry, in many ways they cannot be compared to any of the truly major\nplayers in the field. They do not rank among the large, aggressively\nexpanding fertility businesses that operate dozens of clinics and\nlaboratories worldwide: The largest chains in Finland have only three or\nfour clinics; United States market leader IntegraMed Fertility, for example,\noperates 93 clinics across North America, while the Australian Virtus Health\nruns its 46 clinics not just in Australia, but also in Ireland, Denmark and\nSingapore. The Finnish clinics do not prominently advertise their services\ninternationally, nor are donors or large number of donor gametes imported\nfrom abroad, and donors are not recruited in larger numbers by\n‘overpayments’. Finnish clinics even offer ‘one-stop’ services, unlike\ncompanies that break down the IVF treatment process into stages and then\noutsource some (or all) of those stages to countries with favorable\nregulations and costs. If one wishes to become a big transnational fertility\nservice provider, it seems one has to engage in such market activities\nsooner or later.\n\nIn an IVF lab, embryo culture follows the placing of an egg and sperm into a\npetri dish for fertilization. After fertilization takes place through either\nIVF or intracytoplasmic sperm injection, the resulting embryos are cultured\non petri dishes in one or two different artificial culture mediums until\nthey reach the so-called cleavage stage (days two to three, called the\n‘short culture’ in our fieldwork clinics) or blastocyst stage (days five to\nsix, called the ‘long culture’). After this, a single embryo is transferred\nto the intended mother’s womb. It is a rule of conduct among practitioners\nin Finland, seen in the treatment protocols at many of the clinics, that\nonly one or (in rare cases) two embryos can be transferred simultaneously.\nThe remaining viable embryos are frozen for possible subsequent treatment\ncycles. Sometimes embryos to be used for frozen transfer are also cultured\nafter they have been thawed, to ensure that the embryos are alive and start\ndividing after being frozen.\nDuring the culture period, the embryo petri dishes are kept in artificial\nincubators that mimic the reproductive environment. Traditionally, embryos\nare removed from the incubator for assessment, under a microscope, of their\nquality and stage of development. Instead, the TLS used in some clinics\n(often simultaneously with more traditional, hands-on assessment methods)\ncan take images of embryos at frequent time intervals, which enables their\nassessment without removal from the incubator. A TLS can also apply\n(data-driven) algorithmic software that assists the embryologist to select\nthe best-quality embryo for transfer, potentially improving the chance of a\nlive birth.\n\nDuring our observations, IVF success was often reduced to the issue of embryo\nquality in discussions at clinics and public events. First and foremost, the\nincubator environments were seen as instrumental to ensuring the maintenance\nof embryo quality and even to improving it. This was also discussed at peer\nsupport group meetings. Patients wondered whether their clinics used the\nbest possible incubators, wanted to know about the expenses and whether they\ncould afford the incubator technologies.\nThe incubator technologies used at clinics during our fieldwork included a more\ntraditional incubator from which embryos were removed for assessment on a\ndaily basis, as well as a few different TLS (Vitrolife’s EmbryoScope and\nPrimo Vision, Auxogyn’s Eeva, and Esco Medical’s MIRI). The medical staff\nbelieved that embryo quality could be improved with TLS because the\nexamination was conducted without removing the embryos from the incubator.\nTLS do not otherwise improve embryo quality per se, although some research\n(e.g.  Chen et al.,\n2013 ;  Meseguer et al., 2012 ) suggests that TLS aid the selection of\nthe best-quality embryo for transfer because the embryos are ‘truly’ known\nthrough this technology. Indeed, there is quite a lot of fuss about TLS in\nboth biomedical research and clinical practice. It has been hailed as the\nmost significant and groundbreaking new technology in IVF in decades (see\n Van de Wiel,\n2017 ,  2018 ,  2019 ). Its advocates even believe that TLS will eventually\nmake it possible to recognize chromosomal or genetic deviations in embryos –\nan art that currently requires a biopsy analysis during PGD/PGS (field\nnotes, public lecture at fertility clinic; see also  Daughtry and Chavez, 2018 ).\nThe fuss is apparent in the following snapshot from our field notes from the\n2013 Nordic IVF Laboratory Society conference where a TLS advocate is\ndescribing the technology:  It’s the first day of the Nordic IVF Laboratory Society conference\nand a presentation about the arrival of time-lapse technology is\non. The speaker is an embryologist and is connected to a company\nthat is marketing an application of the device; she hence\ndeclares herself to be a bit ‘biased’ in that regard. At the\nbeginning of the presentation, she shows a colourful cartoon\ndrawing in which embryos drink, smoke and party all night while\nthe biologists are out of the lab. The cartoon shows how the\nbadly behaving embryos hear a biologist returning to work and\npretend to be well-behaved while she has a look at them. After\nshowing the cartoon, the presenter declares pointedly that ‘with\ntime-lapse, the party is over’: Time-lapse technology enables\nstaff to check on video to see how the embryos were behaving\nwhile they were not being monitored. This is important because\nwhen left alone, embryos ‘do funny things’. They are ‘naughty’\nand ‘deceitful’. The speaker explains that with time-lapse ‘we\ncan see how they behave’. Only those that have behaved in a\ncertain way ‘do become babies’ – the naughty ones ‘will not\nbecome babies’.\nIt’s the first day of the Nordic IVF Laboratory Society conference\nand a presentation about the arrival of time-lapse technology is\non. The speaker is an embryologist and is connected to a company\nthat is marketing an application of the device; she hence\ndeclares herself to be a bit ‘biased’ in that regard. At the\nbeginning of the presentation, she shows a colourful cartoon\ndrawing in which embryos drink, smoke and party all night while\nthe biologists are out of the lab. The cartoon shows how the\nbadly behaving embryos hear a biologist returning to work and\npretend to be well-behaved while she has a look at them. After\nshowing the cartoon, the presenter declares pointedly that ‘with\ntime-lapse, the party is over’: Time-lapse technology enables\nstaff to check on video to see how the embryos were behaving\nwhile they were not being monitored. This is important because\nwhen left alone, embryos ‘do funny things’. They are ‘naughty’\nand ‘deceitful’. The speaker explains that with time-lapse ‘we\ncan see how they behave’. Only those that have behaved in a\ncertain way ‘do become babies’ – the naughty ones ‘will not\nbecome babies’.\nThe advocate implies that the rationale for using TLS is that it produces a\n different kind of  knowledge about embryogenesis. TLS\noffers embryologists more visual and temporal information about embryo\ndevelopment during the culture process, as the system constantly monitors\nthe embryo by taking photos every five to twenty minutes. The resulting\nvideos allow the embryologist to observe and record developmental markers,\nsuch as the timing of cell division and the movements of embryo growth. As\nembryologists in our fieldwork clinics told us, they conventionally take ‘a\nquick glance’ manually through the microscope once a day to visually examine\nthe embryo’s static morphological appearance – embryos should not be out of\nthe incubators for more than two or three minutes, or their pH levels will\ndrop. TLS knowledge about embryogenesis, reconstructed by its temporality,\nmay ultimately ‘rewrite the facts of life’, as we discussed with a\nconference exhibitor after the presentation described above. Indeed, one\nmight speak of an epistemic break in the practices of knowing embryos and\ntheir viability (in culture,  in vitro ), as the previously\ninvisible temporal dimension becomes the way of knowing and reproductive\ndecision-making (see  Van de Wiel, 2017 ,  2018 ,  2019 ).\nTLS provide a representation of embryo time which can be manipulated to give\nvisual access to reproductive processes that were previously too slow to be\nobserved.  Van de Wiel\n(2018 : 21) argues that TLS introduce a ‘cinematographic turn’\nin IVF clinics, making embryogenesis visible as a process for a varied\naudience. TLS-based information, videos and images are shared routinely with\nintended parents, allowing patients to visually take part in the scientific\nwork ( Landecker,\n2007 : 123). As  Helosvuori (2019)  has shown,\nwhen embryos are observed through the microscope in the traditional way,\nintended parents are also provided with information about embryo development\nand given a role in decisions on embryo selection. In such cases, parents\nare only  told  about embryogenesis and morphology or shown\n static  pictures of (their) embryos. Nevertheless,\nthrough such information-sharing, patients become enrolled into the process\nof knowledge production and even into interventions that render previous\nconceptions of embryo viability inaccurate. This happens with ‘pity\ntransfers’, where embryos that are (perceived to be) inviable are\ntransferred because the patient wishes it: These transfers sometimes result\nin the birth of healthy babies, which in turn leads embryologists to revise\ntheir views on embryo viability ( Helosvuori, 2019 ).\nAs we hinted at the beginning of this section, patients are aware of TLS and\nits additional costs per treatment cycle. It is also one of the few\nlaboratory technologies directly marketed to patients ( Pottage, 2018 ). Thus, TLS\ninvolves additional clinical and financial decision-making on the patients’\npart.\nIn the following snapshot from a video recording of a clinic appointment, the\nembryologist is showing the intended parents an accelerated video of the\ncultured embryos collated together on-screen:  The embryologist and the intended parent couple are all sitting in\nfront of a big computer screen. The intended mother gives few\nfaint shrieks as the video goes on and points excitedly at the\nevolving embryo images. The embryologist is simultaneously\nexplaining about the two embryos chosen for transfer on the same\nday: ‘We could take all these fertilized oocytes and look at\nthem all together. So let’s start [the video] again. Okay. Let’s\nfollow them [the embryos] here. The first thing that we are\nchecking was these pronuclei here. So, the division with the\noocyte is okay and the normal pronucleus amount [too]. Very\nimportant. And now the next thing that we are looking at in\nthese embryos, which is a good sign for an embryo, when it’s\ndivided on the same day [as] fertilization, and that was on\nSaturday, when you were [at the clinic for the egg-harvesting\nand sperm sample] in the morning. So the [same] afternoon,\naround 25, 26, 27 hours after fertilization, the oocytes start\nto, the pronucleus starts to disappear, like this and then it\ndivides first time [into a two-cell embryo]. So this type of\nembryo is, it’s so-called early [cleavage] embryo. And it’s\nalways when … There has been studies of many embryos in the\nliterature, so it’s always a good sign for an embryo when it\ndivides like this. So, I think now this is going as it should\nbe, divides into two. Here. This is a little bit slower embryo.\n… A little bit slower than this one, but dividing very nicely.\n[Another one of the embryos chosen for transfer is a] little bit\nquicker all the time than this one and now this one follows\nhere, now it’s here, but still with the right timing and, going\nto, you can see this. It’s going nicely. So, there it is at\neight cells and here is eight cells, which is a really, the\nright amount of cells in the embryo on day three’. They talk a\nbit about the third embryo, which will be cultured until the\nblastocyst stage and then possibly vitrified. Then the\nembryologist asks if the couple would like prints of the embryo\nimages. The couple happily accept these and the embryologist\nsays: ‘I will send it to you. This [too], when we see if the\nembryo goes for freezing. So, after that I will send it to you.\nBut then you had a picture from [a baby], going from, so you can\nsee the kind of embryos that have been transferred. So you will,\nI will send it in the email’. (video recording, clinic C)\nThe embryologist and the intended parent couple are all sitting in\nfront of a big computer screen. The intended mother gives few\nfaint shrieks as the video goes on and points excitedly at the\nevolving embryo images. The embryologist is simultaneously\nexplaining about the two embryos chosen for transfer on the same\nday: ‘We could take all these fertilized oocytes and look at\nthem all together. So let’s start [the video] again. Okay. Let’s\nfollow them [the embryos] here. The first thing that we are\nchecking was these pronuclei here. So, the division with the\noocyte is okay and the normal pronucleus amount [too]. Very\nimportant. And now the next thing that we are looking at in\nthese embryos, which is a good sign for an embryo, when it’s\ndivided on the same day [as] fertilization, and that was on\nSaturday, when you were [at the clinic for the egg-harvesting\nand sperm sample] in the morning. So the [same] afternoon,\naround 25, 26, 27 hours after fertilization, the oocytes start\nto, the pronucleus starts to disappear, like this and then it\ndivides first time [into a two-cell embryo]. So this type of\nembryo is, it’s so-called early [cleavage] embryo. And it’s\nalways when … There has been studies of many embryos in the\nliterature, so it’s always a good sign for an embryo when it\ndivides like this. So, I think now this is going as it should\nbe, divides into two. Here. This is a little bit slower embryo.\n… A little bit slower than this one, but dividing very nicely.\n[Another one of the embryos chosen for transfer is a] little bit\nquicker all the time than this one and now this one follows\nhere, now it’s here, but still with the right timing and, going\nto, you can see this. It’s going nicely. So, there it is at\neight cells and here is eight cells, which is a really, the\nright amount of cells in the embryo on day three’. They talk a\nbit about the third embryo, which will be cultured until the\nblastocyst stage and then possibly vitrified. Then the\nembryologist asks if the couple would like prints of the embryo\nimages. The couple happily accept these and the embryologist\nsays: ‘I will send it to you. This [too], when we see if the\nembryo goes for freezing. So, after that I will send it to you.\nBut then you had a picture from [a baby], going from, so you can\nsee the kind of embryos that have been transferred. So you will,\nI will send it in the email’. (video recording, clinic C)\nNote that TLS seems to fulfill both an affective and a clinical diagnostic\nfunction here. The embryologist straightforwardly juxtaposes pictures of\nviable embryos with baby pictures, not unlike ultrasound images and videos.\nAs with ultrasound, prenatal life in TLS imaging videos is known without the\nembodied, experience-based knowledge of the intended mother (see e.g.  Duden, 1993 ;\n Homanen,\n2013 ). However, unlike ultrasound, TLS places the prenatal life\nconcretely outside the maternal body and rewrites the human origin story\naccordingly ( Van de\nWiel, 2017 ,  2018 ). Nevertheless,\nencountering prenatal life – seeing ‘the baby’ for the first time and\ngetting a first picture – is the obvious attraction for the intended\nparents.\nThe embryologist in the snapshot also seems to be promoting the notion that\nthere is a universal regularity in the temporal process of embryo\ndevelopment that can be both observed and operationalized to\n predict  viability (see  Waldby, 2019 ). Here, she refers\nto studies of embryo development, but TLS also have a data-driven component.\nThis component allows visual information about temporal and morphological\nfeatures (which are sometimes invisible to the human eye) to be measured,\nquantified and analyzed through algorithms to predict embryo viability.\nThe algorithms are based on historical data sets: The timing of cellular\ndivisions is viewed in light of historical embryo populations to predict\nsuccess rates, that is, the likelihood that the treatment will result in\npregnancy. Clinics are also encouraged to provide their own data for the\nfurther development of the algorithmic tools. In principle, then, TLS allow\nclinics to adopt a quantified, automated, standardized, data-driven method\nfor knowing about viability and selecting embryos through prediction (see\n Van de Wiel,\n2018 ,  2019 ).\nWhere TLS run automated algorithms to predict the viability of embryos cultured\nat a clinic, one should ideally be able to use this to scale up production\nby standardizing one’s knowledge production. Data-based prediction is\nmarketed as making it possible to transfer embryos at an earlier stage than\nwould normally be the case, allowing clinics to accommodate more patients\nwith a better success rate.\nDuring our fieldwork, not all clinics took full or even any advantage of this\ndata-driven component, and many embryologists relied at least partially on\nmanual appraisals of embryos, either under the microscope or by rewinding\nTLS videos backwards and forwards on the computer screen. Embryologists\ndescribed how one develops ‘an eye’ and ‘a feel’ for embryos after observing\nthem for a long time. This experience-based, hands-on method of examining\nembryos (i.e. craft know-how) was called ‘a natural way to rule out embryos’\nby one of our informants. Some, however, downplayed their experience-based\nobservational skills in embryo examination, regarding it as pointless in\ncomparison with standard quantified information on embryos. This shows that\nembryologists work at the crossroads of standardized and tacit modes of\nknowledge production, and that technologies can be used in non-standard and\nnon-automated ways.\nDespite TLS’s acclaimed potential to produce valuable information about\nembryos, its translation into clinical use is not clear-cut. To take full\nadvantage of the technology, staff training and a lot of extra time in the\nlaboratory would be required. ‘One could spend half a day tinkering with\nthem [TLS]’, laughed one embryologist in our study. On the other hand, the\nmeaning of all the markers of embryo development on which TLS reports is not\nfully known. For instance, we were told that in light of historical data\nsets, it is believed that cells should divide in a ‘disjunct manner’ so that\nthe cell ‘is not doing something all the time’. However, when we asked\nfurther about the meaning of this, we were told that there is no knowledge\nabout what it specifically means for cell quality, let alone for further\nembryo development.\n\nKnowing developmental time in the IVF process (through predictive TLS in\nincubators or by manual e/valuation) affects the decisions on the duration\nof embryo culture in incubators: For how many days after fertilization\nshould the lab culture embryos that are candidates for transferral to an\nintended mother’s uterus? There seems to be no strong consensus among\nclinicians about which transfer stage results in more pregnancies or about\nthe health implications of this for IVF-born children. Clinics mostly follow\ntheir own experience-based assessments and small-scale statistics and\npractice accordingly. Thus, the duration of embryo culture varies, leading\nto different stages of embryogenesis at transfer.\nThe embryologist places the eggs and sperm on a petri dish on the day of\novarian pick-up (OPU), that is, the day when eggs are harvested from the\novaries. Two preliminary nuclei (one from the egg and one from the sperm)\nsignify fertilization one day after OPU (day one). Markers indicate whether\nthe cells have begun cleavage (i.e. to form an embryo) two days after OPU\n(day two). There are six to twelve cells on day three; the morula stage is\nreached on day four. A blastocyst forms ‘roughly on time’ by days five to\nsix.\nThe main stages at which embryo transfer is performed are the cleavage stage\n(days two to three after fertilization) and the blastocyst stage (days five,\nsix or even seven after fertilization). Some of our field clinics advocated\nthe ‘long culture’ process of five to six days and some preferred the ‘short\nculture’ of two to three days. Some even changed their preferences and\npractices during our fieldwork period. According to our informants, public\nuniversity hospitals routinely used the short culture to accommodate more\npatients and save resources.\nIn our fieldwork clinics, the short culture was argued for in terms of economy\nand the superiority of the ‘natural’ womb environment. The economic\narguments referred not to the cost-effectiveness of treating as many\npatients as possible, as fast as possible, with the fewest resources (even\nthough this might also have been taking place), but rather to being\neconomical in the face of the chronic uncertainty of prediction that\ncharacterizes the IVF treatment sector. One never knows if embryos that look\nviable on days one to three will survive until the blastocyst phase.\nEspecially with poor-quality embryos, medical staff were keen on cleavage\nstage transfer ‘so that there is at least something to transfer’, as one\nembryologist put it. These arguments were based on the belief that the womb\nenvironment is always better for embryo development than the artificial\nincubator and growth medium, placing hope in the natural(ized)  in\nvivo  environment to enable viability. This belief was further\nunderpinned by the uncertainty regarding whether embryos that are judged to\nbe of good quality (according to morphology and developmental biomarkers)\nreally implant more often than bad-quality embryos (see  Thompson, 2005 :\n114). Some professionals at laboratories with TLS also relied on the\nsystem’s predictive analysis and standardized use: What need is there for\nthe long culture if one can predict longer-term viability?\nThe prediction of viability is also questioned by professionals. Sometimes\nmarkers that appear to be unpromising turn out to be meaningless in a long\nculture, again because of uncertainty regarding the meaning and relevance of\nall the stages in development and morphological characteristics. This is\napparent in the following interview extract:  Q: Could you explain once more what you observe from the\nembryos? A: So we observe in day two and three the cell quantity, … how much\nfragmentation there is [a sort of graininess on the embryo] and\nthen we have a look at the cells, whether they are the same or\ndifferent size, it’s better if they are similar size. … Then we\nhave a look at the nuclei, is there a nucleus in every cell. If\nthere are multiple nuclei, that is a bad situation, then there\nhas been something wrong in the cleavage. Q: But that can be fixed later on, is that right? A: Yes, that can be fixed. We used to be much more critical toward\nmultiple nuclei and so we dumped those always, mostly, when we\nsaw even one cell with multiple nuclei. But now we have noticed\nthat when they are transferred, it can get fixed, so it is not\nsuch a critical factor after all. (embryologist 1, clinic A)\nQ: Could you explain once more what you observe from the\nembryos?\nA: So we observe in day two and three the cell quantity, … how much\nfragmentation there is [a sort of graininess on the embryo] and\nthen we have a look at the cells, whether they are the same or\ndifferent size, it’s better if they are similar size. … Then we\nhave a look at the nuclei, is there a nucleus in every cell. If\nthere are multiple nuclei, that is a bad situation, then there\nhas been something wrong in the cleavage.\nQ: But that can be fixed later on, is that right?\nA: Yes, that can be fixed. We used to be much more critical toward\nmultiple nuclei and so we dumped those always, mostly, when we\nsaw even one cell with multiple nuclei. But now we have noticed\nthat when they are transferred, it can get fixed, so it is not\nsuch a critical factor after all. (embryologist 1, clinic A)\nEmbryos may develop and implant normally further along the line. There are also\ngenome abnormalities that cannot be predicted in early-term cultures.\nRecurrent early miscarriage, for example, may be a sign of problems with the\nsperm genome: It is believed that the genome in the sperm is only activated\nin the embryo after day three. Problems will therefore not manifest\nthemselves during the cleavage stage and cannot be predicted without\nPGS/PGD.\nBoth of the above cases are also good examples of problems not just with\nprediction in general, but with TLS prediction in particular. TLS analyses\nare not (yet) sensitive to most chromosomal abnormalities, nor do they take\nnew or local conditions into account very well. This is because they are\nbased on  historical  data sets and not all clinics enter\ntheir own embryo population data into the data pool ( Van de Wiel, 2018 ).\nMoreover, in contrast to the hype around TLS, some clinics have almost stopped\nusing it. Long culture advocates at some of our fieldwork clinics told us\nthat they did not need the (far-from-foolproof) predictive component of TLS\nbecause it did not really benefit clinical IVF outcomes unless one was set\non doing only short cultures. Here, TLS was described as ‘merely a good\nincubator’ and thus as too expensive for clients.\nThe head doctor explained to me in his office: ‘It is of course\ninteresting to see the temporal development [on-screen] and find\nout exactly how the cell can fix itself, but it does not matter\n[that much] clinically what the journey has been like until\nblastocyst if it looks pretty. … When you master the long\nculture [manually] in the laboratory, you don’t need TLS\nanymore. It is only useful when you don’t have the skill to keep\nthe embryos alive [until blastocyst]. … This idea of the womb\nalways being like a warm embrace for the embryo is just not true\nbecause the womb is better on day five or six’ [than days two to\nthree, referring to the endometrium being more receptive at that\nstage of the IVF cycle]. He then went on to elaborate on how\nthere are ways to test the stage of the endometrium to define\nthe exact individual implementation window for the transfer and\nhow that combined with PGS/PGD is the way to IVF success. (head\ndoctor, clinic F)\nThe head doctor explained to me in his office: ‘It is of course\ninteresting to see the temporal development [on-screen] and find\nout exactly how the cell can fix itself, but it does not matter\n[that much] clinically what the journey has been like until\nblastocyst if it looks pretty. … When you master the long\nculture [manually] in the laboratory, you don’t need TLS\nanymore. It is only useful when you don’t have the skill to keep\nthe embryos alive [until blastocyst]. … This idea of the womb\nalways being like a warm embrace for the embryo is just not true\nbecause the womb is better on day five or six’ [than days two to\nthree, referring to the endometrium being more receptive at that\nstage of the IVF cycle]. He then went on to elaborate on how\nthere are ways to test the stage of the endometrium to define\nthe exact individual implementation window for the transfer and\nhow that combined with PGS/PGD is the way to IVF success. (head\ndoctor, clinic F)\nThis head doctor sketches a wholly different picture of TLS as a clinical tool\ncompared with industry advocates or with what the technology’s worldwide\nsales figures might suggest (e.g.  Van de Wiel, 2019 ). Instead of\ncelebrating detailed knowledge of embryogenesis for clinical purposes, he is\nof the opinion that TLS is rendered redundant by skilled laboratory work\nlearned through repeated performance over time. He does not share the\nconception that embryo transfer in the blastocyst phase can be justified\nwith arguments about the caring  in vivo  uterine environment\nbecause the optimal time for implantation is understood to be around days\nfive to six, with some individual variation.\nLong culture is also perceived as economic, but in somewhat opposite ways to\nshort culture. When long culture is mastered, it is believed, viable embryos\nare likely to survive. This saves money and labor for all:  It really is a lot more cost-effective, in a way, if we continue\n[to culture] to day five because many times it can occur that in\nday two, there are a lot of embryos, and then you transfer one\nof them into the uterus, and many into the freezer, and then you\ntransfer those day two embryos [later]. It can happen that the\npatient undergoes multiple transfers with frozen embryos and not\nwith great odds. However, if you continue to day five, you sort\nof weed out useless [non-viable] embryos … and then the odds per\ntransfer are much higher. (doctor 2, clinic A)\nIt really is a lot more cost-effective, in a way, if we continue\n[to culture] to day five because many times it can occur that in\nday two, there are a lot of embryos, and then you transfer one\nof them into the uterus, and many into the freezer, and then you\ntransfer those day two embryos [later]. It can happen that the\npatient undergoes multiple transfers with frozen embryos and not\nwith great odds. However, if you continue to day five, you sort\nof weed out useless [non-viable] embryos … and then the odds per\ntransfer are much higher. (doctor 2, clinic A)\n‘Cost-effectiveness’ refers to both patients’ and clinics’ interests. The\nprofessionals explained to us that transferring as many embryos as possible\nwith the least effort is not good for business at the end of the day. The\npregnancy results remain low and freezing multiple embryos also takes time\nand effort. Patients are burdened with repeated disappointments and\nunnecessary medical interventions. It is more lucrative to culture for\nlonger and to transfer and freeze less. According to this industry logic,\ngood news will eventually travel, meaning that in addition to better\nclinical IVF outcomes, business outcomes improve too.\nThe business logic, then, does not exclude care for patients or cells, but\nrequires it. Patients’ wishes and practical everyday lives affect the\nculture duration in other ways as well. Craft is about practicalities (e.g.\n Meskus,\n2018 ). Cleavage stage transfers may be made for patients who\nwant their transfer to be conducted by a particular doctor who can only\nperform the transfer on the second day of embryo culture. National holidays\nare often accommodated. Furthermore, if a patient is eager to ‘get to the\npoint of embryo transfer’ – to ‘push the panic button’, as put in the words\nof one informant (doctor 2, clinic A) – rather than waiting a couple of days\nand incurring the risk that the embryos might perish by days five to six, a\ntransfer is made to please them.\nOverall, TLS are designed to enable an automated, standard way of knowing\nembryo viability in terms of embryogenesis and morphology and are thus\ndesigned to enable the scaling-up of production. As with standards and\nstandard technologies more generally, they are subject to local adjustment\nand manipulation in practice ( Timmermans and Epstein, 2010 ).\nThe unpredictability, uncertainty, locality and individuality of embryo\nviability make it hard for the TLS standard to work and for clinics to\ncapitalize on it. Furthermore, there are practices that resist this\nautomation and standardization for its own sake, for the sake of the\nintended parents’ lives and finances as well as for those of the clinics.\nHere, craft has both economic and ethical value.\n\nThe artificial embryo growth medium in which the fertilized ovum is immersed\nduring culture contains mostly glucose, pyruvate and energy-providing\ncomponents. It is also possible to add amino acids, nucleotides, vitamins\nand cholesterol to improve embryo growth and development. All in all, the\nmediums aim to mimic the optimal reproductive environment for each\ndevelopmental stage (see  Landecker, 2016 ).\nHowever, the big pharmaceutical companies that provide the mediums do not share\nall the details, such as add-ons or percentages of different chemical\ningredients in their products, despite pressure from medical practitioners.\nFor instance, according to our informants, the Nordic Fertility Society has\ntried to force companies to reveal this information without success. The\nmedical staff we talked with told us that in the early days of clinical IVF\nwork, they had been able to make their own mediums as laboratory craftwork;\nnow they were dependent on standard mediums provided by large companies,\nbecause EU regulations only permitted CE-tested and certified mediums, which\nonly big companies were able to produce.\nAs a result, the choice of a medium for one’s IVF laboratory is made without\nfull information.\nWe are dependent on the commercial growth mediums [rather than\nbeing able to make them in-house]. The one thing I have always\nwished for is that at some point the producers of the mediums\nshould be supervised, just like IVF medications are controlled.\nThey are not controlled. They still have these product secrets\nabout what gets added into the mediums. I really hope there will\nbe [a change] because we put embryos in them. And these firms\ncan add some preparation that will make more beautiful embryos,\nembryos that develop faster. What are the long-term effects?\nOverall, I think we should pay attention not so much to what\nkind of results the clinic has, how many pregnancies we manage\nto induce, but think more about the effects [of embryo culture\nand IVF] on the individual’s health and the health of the\nchildren and health later in life. This is something that has\nalways worried me personally. … What is being done when, for\ninstance, some growth hormone is added into some mediums. No one\nknows what effects it has. And then it may be a commercial\nsecret, so we don’t even know what exactly is in them.\n(embryologist 4, clinic C)\nWe are dependent on the commercial growth mediums [rather than\nbeing able to make them in-house]. The one thing I have always\nwished for is that at some point the producers of the mediums\nshould be supervised, just like IVF medications are controlled.\nThey are not controlled. They still have these product secrets\nabout what gets added into the mediums. I really hope there will\nbe [a change] because we put embryos in them. And these firms\ncan add some preparation that will make more beautiful embryos,\nembryos that develop faster. What are the long-term effects?\nOverall, I think we should pay attention not so much to what\nkind of results the clinic has, how many pregnancies we manage\nto induce, but think more about the effects [of embryo culture\nand IVF] on the individual’s health and the health of the\nchildren and health later in life. This is something that has\nalways worried me personally. … What is being done when, for\ninstance, some growth hormone is added into some mediums. No one\nknows what effects it has. And then it may be a commercial\nsecret, so we don’t even know what exactly is in them.\n(embryologist 4, clinic C)\nThis embryologist sees the fact that commercial establishments often do not\ndisclose the exact composition (see  Landecker, 2016 ) of their\nmediums as ethically dubious and as needing to be controlled in the way that\nIVF medications are controlled. There is suspicion that pharmaceutical\ncompanies add chemicals to the mediums to make the embryos morphologically\nbeautiful and enhance their developmental performance – to improve the\nembryo vitality as assessed in the grading models that ultimately lead to\nthe choice of an embryo for transfer. There is concern about the\nimplications of this for the long-term health of embryos and children.\nThis concern seems legitimate, as can be seen from a story one embryologist\ntold us about a sudden decrease in successful IVFs at her clinic: ‘The\nmedium manufacturer had changed something. That was revealed, but first they\nsaid they had made no changes, but then there came news from other places as\nwell that there have been problems. Then they confessed’ (embryologist 1,\nclinic E). It seems that the absence of information from the pharmaceutical\nindustry, which is explained away in terms of market advantage, not only\nhinders the possibility of producing and knowing embryonic human life, but\nis also a liability for that life  in vitro .\nEmbryologists are aware that culture mediums only partially and artificially\nmimic uterine surroundings, turning optimization into an uncertain and risky\naffair. In contrast to the historical institutional goal of neutralizing\nvariability and making environments inert, the medium is reconceptualized\nfrom a mere adequate background condition to a constitutive factor in the\nmaking and knowing of human life forms. This scientific approach to\nepigenetics necessitates attention not just to the medium’s biomaterials as\ninteractive agents, but also to the social, cultural, economic and political\nconstitution of the material setting ( Alder, 2013 ;  Landecker, 2016 :\n149).\nThe industrial setting of culture mediums for IVF also means that individual\nclinics and chains do not share knowledge about their materials and\npractices for business reasons. This caused frustration among the\nembryologists in our study:  Everyone does [embryo culture] in their own different ways [refers\nto clinics] and it annoys me enormously that in this business\nresults are not comparable, really. You just think that, oh, so\nyou are transferring that sort of an embryo, I wonder what it\nwould look like at our place [if we were to choose same embryo].\nIt is an established fact that embryos grow faster in some\ngrowth mediums. If you think about that, when you examine some\nother [embryo], it looks at this moment like that. Can you trust\nanything? If everything has an effect, it makes you feel awful,\nlike this is not working. There is no absolute [truth] when the\nreality is that it varies. … But then we think, shall we use a\ndifferent medium, but how do we know [how they work]? Based on\none medium we could say this is a good embryo and based on\nanother we could say this is too fast and this is bad.\n(embryologist 2, clinic C)\nEveryone does [embryo culture] in their own different ways [refers\nto clinics] and it annoys me enormously that in this business\nresults are not comparable, really. You just think that, oh, so\nyou are transferring that sort of an embryo, I wonder what it\nwould look like at our place [if we were to choose same embryo].\nIt is an established fact that embryos grow faster in some\ngrowth mediums. If you think about that, when you examine some\nother [embryo], it looks at this moment like that. Can you trust\nanything? If everything has an effect, it makes you feel awful,\nlike this is not working. There is no absolute [truth] when the\nreality is that it varies. … But then we think, shall we use a\ndifferent medium, but how do we know [how they work]? Based on\none medium we could say this is a good embryo and based on\nanother we could say this is too fast and this is bad.\n(embryologist 2, clinic C)\nThis embryologist is very aware that the embryos as known are the product of\nthe growth medium. This knowledge also has consequences: The embryos are\ngraded differently and different kinds of embryo are transferred into the\nintended mothers’ wombs at different clinics. However, as with the large\npharmaceutical corporations, it is not in the local private clinics’ market\ninterest to share their data: They prefer to keep their best practice,\nacquired through time-consuming laboratory tinkering and trial and error, to\nthemselves. There is economic value in knowledge that is not shared. In the\ncommercial industry setting, economic value wins out over the ethical value\nof openness.\nThe valuable knowledge here is not just knowledge of the type of medium used,\nbut also local know-how (see  Levin and Leonelli, 2017 ).\nMaking a standard medium work is not automatic. Rather, it is an\naccomplishment arrived at in the laboratory through careful, non-standard\ncraftwork. The embryologists in our study seemed reluctant to change the\nmedium brand they were using, which they had found to  work \nby doing and experiencing laboratory work.\nWe don’t change them often. We start using [new mediums] if … it is\nmore usable, like easier to use. (embryologist 1, clinic E) Sometimes we test new mediums from a different company. Usually I\nthink there is no difference. These tools are often [chosen]\naccording to what fits your hand the best and which one helps\nyou work better [with the embryos]. (embryologist 2, clinic\nD)\nWe don’t change them often. We start using [new mediums] if … it is\nmore usable, like easier to use. (embryologist 1, clinic E)\nSometimes we test new mediums from a different company. Usually I\nthink there is no difference. These tools are often [chosen]\naccording to what fits your hand the best and which one helps\nyou work better [with the embryos]. (embryologist 2, clinic\nD)\nThe industrially produced (and not fully known) mediums are tested on embryos\nto see how they work in practice, to see how the embryos adapt to them. Via\ntheir own testing, the clinics achieve higher pregnancy rates and improve\nthe practicalities of everyday laboratory work. Making mediums work by\ntesting involves not just the context of production of the medium itself,\nbut also the individual clinic and its hands-on labor. However, in clinical\nIVF settings, the professionals are not able to isolate the substances that\nare necessary or harmful for continued life because the composition of the\nmediums is not known. The professionals also don’t know the (probably)\ninnumerable empirical tinkerings and tests that take place in pharmaceutical\nindustry laboratories before a medium is released to market, even though\nthose tests are part of the co-constitution of future embryos, inseparable\nfrom their  in vitro  milieu. As one embryologist summed up,\n‘we just have to trust that they have been properly tested’ (embryologist 2,\nclinic D).\nFor the same reason of non-disclosure, many embryologists prefer to choose all\ntheir mediums from the same provider – the embryo growth medium, freezing\nmedium, thawing medium and so on – as if they will fit together better. For\npharmaceutical companies, this purchasing logic means that selling one\nmedium equals selling a whole product line.\nWe choose the manufacturers according to experience and then we\nprefer to take all of the mediums from the same manufacturer\nbecause they work together well …. It is difficult to change\nthem because they have different substances in them and we want\nthe whole family then, so to speak. (embryologist 4, clinic\nC)\nWe choose the manufacturers according to experience and then we\nprefer to take all of the mediums from the same manufacturer\nbecause they work together well …. It is difficult to change\nthem because they have different substances in them and we want\nthe whole family then, so to speak. (embryologist 4, clinic\nC)\nOne might easily think that standard medium products bought from big\ntransnational pharmaceutical providers would make business easier for local\nclinics, as they would not have to invest time and resources in making their\nown mediums. Although it may take a period of careful craftwork to get\npurchased mediums to work with the embryos, the clinics should be able to\nstick with the same products for a long time thereafter, without having to\nrepeat the process. This industrial logic of cost-effectiveness is certainly\nsomething the big pharmaceutical companies promote. However, in some\nrespects, both economic and ethical value gets lost in these forced dealings\nwith pharmaceutical corporations, since information on embryo epigenetics is\nnot shared.\nIt is no surprise, then, that when talking about mediums with the professionals\ntoday, they bring into the conversations the early days of maverick small\nbusinesses, when they used to craft their own. Such practice appears as a\n(politically) more desirable practice, with a different distribution of\nknowledge.\n\nIn this article, we have explored embryo culture in clinical IVF laboratories\nas a knowledge production practice and process that ultimately aims to know\nand select the most viable embryos for transfer. We have discussed the ways\nin which incubator technologies (with and without TLS add-ons) and embryo\nculture mediums enable and disable knowledge about embryo morphology,\nembryogenesis and epigenetics. Our results regarding knowledge production\nare a contribution to discussions of craftwork and standardization in\nbioindustrialisation, which have hitherto mostly been discussed in the\ncontext of research laboratories rather than clinical laboratories (however,\nsee e.g.  Pavone and\nArias, 2012 ;  Van de Wiel, 2019 ).\nPrior research suggests that transnational pharmaceutical and biotechnological\ngiants rule the market for the commercialized products used in IVF; they\nalso aggressively promote standardization and automation in clinics ( Franklin, 2013 ;\n Global Fertility\nAlliance, 2018 ;  Van de Wiel, 2019 ).\nStandardization and automation can be also seen as requiring clinics to\nscale up and branch out into new areas, geographically and otherwise ( Franklin, 2013 ;\nsee also  Meskus,\n2018 ). Despite this interest in capitalizing on the expanding\nfertility market and life science-led expansions more broadly, automation\nand standardization are only possible up to a point. This means that the\nfuture of bioindustrialization depends on the successful financializaton and\ncommercialization of a mixture of human effort (patients and medics, in our\ncase), standards, machinery and the labor of living cells, as  Franklin (2013) \npoints out. Our study offers a novel view into a few such possible mixtures\nin IVF laboratories.\nOur study confirms previous findings ( Foley and Whitaker, 2012 ;  Muniesa et al.,\n2017 ;  Petersen and Krisjansen, 2015 ) that there is a persistent gap\nbetween the enthusiastic market expectations and fuss around new\nbiotechnologies and their actual industrial success, including those in\nclinical laboratories. TLS is a good example of this. While it is celebrated\nas a revolutionary technology that makes an epistemic break in the practices\nof knowing embryo viability by enabling staff to predict development times\n(see also  Van de Wiel,\n2018 ,  2019 ), TLS turns out to be not so practical in clinical\npractice after all, or gets used in non-standard and non-automated craft\nways. This is not just because of the uncertainties and unpredictabilities\nof the performance of living material, but also because local economic\ninterests do not always coincide with the larger industrial interests in\nautomation and do not exclude concern over patients’ finances.\nFurthermore, while it is perhaps interesting for research, much of the detailed\nknowledge about morphological features and embryogenesis enabled by TLS\nseems to make no difference in clinical work and the predictive component –\nwhich is based on algorithmic analyses of historical data sets and ideally\nallows the automation of knowledge production – is regarded as unnecessary\nand lacking. The craft of embryo culture also has value in itself: The\nvalued goal is to keep embryos perceived as optimal for transfer alive and\nwell  in vitro  until the blastocyst phase.\nThe second technology of much embryo culture – embryo culture mediums – enables\nand enacts knowledge about not just embryogenesis, but also epigenetics.\nHowever, because of supranational and national safety requirements,\npharmaceutical giants have a monopoly over the mediums’ distribution and\nthey refuse to share information about their composition. This is not\nuncommon in commercial settings ( Landecker, 2007 ,  2016 ). In order\nto ‘get the job done’, clinicians simply need to trust the standards (see\n Timmermans,\n2015 : 80). However, professionals call for control from\nregulators to force this information into the open in the name of liability\nfor human life.\nArguably, in our view, this lack of regulation results from the historical\nunderstanding of mediums as mere uterine-like backgrounds, rather than as\nco-constitutive artificial agents in embryo culture, which is how mediums\nare perceived in epigenetics. This suggests that political attention to the\nissue is needed. After all, the ultimate ‘product’ of the laboratory\npractices in IVF is a human person and the practices might bear consequences\nfor following generations.\nWhile commercial clinical establishments are not under the same pressure as\nscientific research to share knowledge/data ( Birch, 2017 ;  Levin and Leonelli,\n2017 ), they are embedded in multiple exchanges and\nexpectations.  Hilgartner (2012)  has argued that data-sharing follows a\n‘dialectic of revelation and concealment’ through which knowledge is\nstrategically made available and unavailable. In the case of IVF clinics,\nwhat happens in a situation where there is no open information about\nstandard medium composition is that clinics make standard commercial mediums\nwork in their own laboratories, which involves bioassay experimentation with\nembryos to see how they adapt to it. This experimentation is time-,\nresource- and labor-intensive and involves and develops local know-how.\nThus, it comes as no surprise that the clinics, in turn, do not want to\nshare this valuable know-how with other clinics or the larger industry: It\nis an achievement they want to capitalize on themselves. However, this\nresults in less knowledge about the mediums’ epigenetic effects, further\nconsolidating the bioindustry market, since embryologists feel that it is\nrisky not to buy a whole line of laboratory pharmaceutics from the same\nprovider.\nThe non-disclosure of milieu information, which affect embryonic life, causes\nfrustration among professionals, who feel responsible toward embryos, future\nbabies and intended parents. They are emotionally invested in the craftwork,\nand as part of this, disclosure/non-disclosure involves ethical as well as\nmarket value. Laboratory work is about both instrumentality and care, which,\nas  Meskus (2018) \nargues, are mutually inclusive and interdependent when one seeks to make\nbiomaterial work in the ways hoped and planned in the age of\nbioindustrialization (see also  Adamson, 2010 ;  Davies and Horst,\n2015 ).\nCan we then say that there are multiple bio-economies being drafted in clinical\npractices, albeit not very explicitly all the time? These also include\nbio-economies where mass production is not the ultimate goal – that is,\nwhere the primary obstacles to automation and standardization are not\nbiological uniqueness, uncertainty or unpredictability. Rather, we are\ntalking about practices where craft kicks back. Such practices can also be\nseen as models for reproductive care, drawing on a value system that\ndisdains the mass industrial approach.","source_license":"CC-BY-4.0","license_restricted":false}