Technology and Pregnancy
2020; Mary Ann Liebert, Inc.; Volume: 22; Issue: S1 Linguagem: Inglês
10.1089/dia.2020.2506
ISSN1557-8593
AutoresJennifer M. Yamamoto, Helen Murphy,
Tópico(s)Diabetes and associated disorders
ResumoDiabetes Technology & TherapeuticsVol. 22, No. S1 Original ArticlesOpen AccessTechnology and PregnancyJennifer M. Yamamoto and Helen R. MurphyJennifer M. YamamotoDepartments of Medicine and Obstetrics and Gynecology, Cumming School of Medicine, University of Calgary, Calgary, Alberta, CanadaAlberta Children's Hospital Research Institute, Calgary, CanadaSearch for more papers by this author and Helen R. MurphyCambridge University Hospitals NHS Foundation Trust, Cambridge, UKWomen's Health Academic Centre, Division of Women's and Children's Health, King's College London, London, UKNorwich Medical School, Floor 2, Bob Champion Research and Education Building, James Watson Road, University of East Anglia, Norwich Research Park, Norwich, UKSearch for more papers by this authorPublished Online:18 Feb 2020https://doi.org/10.1089/dia.2020.2506AboutSectionsView articleView PDFView PDF Plus ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail View articleIntroductionThe manuscripts chosen for this year's technology and pregnancy article provide new insights into fetal exposure to maternal glucose during pregnancies complicated by gestational diabetes, and type 1 diabetes. The increasing use of continuous glucose monitoring (CGM) both in research studies and in real-world settings provide data relating maternal glucose profiles to neonatal health outcomes. These data confirm the well-recognized gestational limitations of glycated hemoglobin (HbA1c), which is inadequate for assessing short-term changes in maternal glycemia during pregnancy.Data from Sweden confirm that we are far from achieving the stringent international consensus target of ≥70% time in range (TIR) 63–140 mg/dL during type 1 diabetes pregnancy (1). Kristensen et al. demonstrate that despite 70% of women achieving HbA1c 140 mg/dL). Taken together, these data suggest that healthcare providers should focus on increasing CGM TIR and decreasing TAR in small achievable increments.Another CONCEPTT ancillary study demonstrated that insulin pump users had a 5% lower (TIR 63–140 mg/dL 53 vs 48%) at 24 weeks, despite comparable first and third trimester CGM glucose measures (4). These findings suggest that clinicians and insulin pump users need to be more aggressive with their insulin dose increments, as gestational insulin resistance increases from around 18–20 weeks (5). Data from the T1D Exchange clinic registry confirms increasing diabetes technology use, with 74% of recently pregnant women using insulin pumps and 36% using CGM (6). Despite increased diabetes technology use, glucose control and pregnancy outcomes remained suboptimal, with LGA rates of 65%. Given the high rate of insulin pump use before and during pregnancy, more data are needed to inform optimal implementation of insulin pump therapy, especially during the second and third trimesters.Current recommendations advise women with gestational diabetes mellitus (GDM) to perform glucose testing at least four times daily, with increasing use of CGM expected as sensors become more affordable, accurate, and user friendly. Using functional data analysis, Law et al. demonstrated that a higher mean glucose level is driven by suboptimal nocturnal glucose control and that small differences in nocturnal CGM measures are associated with offspring LGA (7). Another team used spectral clustering approaches to develop a new glucose variability metric, the “gluocotype” (8). Hall et al. describe three glucotypes of increasing variability (low, moderate, and severe), which together with mean CGM glucose explain >70% of temporal glucose variability. This methodology could detect earlier preclinical forms of dysglycemia and identify those most at risk for type 2 diabetes or prediabetes, which is very relevant for women with GDM. Interestingly it also highlighted the imitations of the oral glucose tolerance test (OGTT), which was seemingly normal in one-quarter of participants categorized as having severe glucotypes. Retnakaran et al. describe the impact of higher environmental temperatures on OGTT results, although unfortunately lacked CGM data, to establish whether or not environmental temperatures impact on CGM measures (9).The follow-up study of over 4,700 mother–child pairs from the Hyperglycemic and Adverse Pregnancy Outcomes (HAPO) study confirmed that, women with GDM were more likely to develop type 2 diabetes or prediabetes (52.2 vs 20.1%) than those in the general maternity population (10). Another study not directly involving pregnant women with diabetes but very relevant to maternity and pediatric populations describes the long-term impact of being born LGA (11). In a large German cohort of over 50,000 participants, almost half of all LGA newborns continued to be overweight or obese into adolescence. This highlights the importance of optimizing maternal glucose control in the second and third trimesters to reduce neonatal LGA and the longer-term consequences of overweight and obesity persisting into adolescence.Key Articles Reviewed for the ArticleContinuous glucose monitoring in pregnant women with type 1 diabetes: an observational cohort study of 186 pregnanciesKristensen K, Ögge LE, Sengpiel V, Kjölhede K, Dotevall A, Elfvin A, Knop FK, Wiberg N, Katsarou A, Shaat N, Kristensen L, Berntorp KDiabetologia 2019;62: 1143–1153Maternal glycaemic control and risk of neonatal hypoglycaemia in type 1 diabetes pregnancy: a secondary analysis of the CONCEPTT trialYamamoto JM, Corcoy R, Donovan LE, Stewart ZA, Tomlinson G, Beardsall K, Feig DS, Murphy HR; on behalf of the CONCEPTT Collaborative GroupDiabet Med 2019;36: 1046–1053Pumps or multiple daily injections in pregnancy involving type 1 diabetes: a prespecified analysis of the CONCEPTT randomized trialFeig DS, Corcoy R, Donovan LE, Murphy KE, Barrett JFR, Sanchez JJ, Wysocki T, Ruedy K, Kollman C, Tomlinson G, Murphy HR; on behalf of the CONCEPTT Collaborative GroupDiabetes Care 2018;41: 2471–2479Diabetes technology use among pregnant and nonpregnant women with T1D in the T1D ExchangePolsky S, Wu M, Bode BW, DuBose SN, Goland RS, Maahs DM, Foster NC, Peters AL, Levy CJ, Shah VN, Beck RWDiabetes Technol Ther 2018;20: 517–523Suboptimal nocturnal glucose control is associated with large for gestational age in treated gestational diabetes mellitusLaw GR, Alnaji A, Alrefaii L, Endersby D, Cartland SJ, Gilbey SG, Jennings PE, Murphy HR, Scott EMDiabetes Care 2019;42: 810–815Association of gestational diabetes with maternal disorders of glucose metabolism and childhood adiposityLowe WL Jr, Scholtens DM, Lowe LP, Kuang A, Nodzenski M, Talbot O, Catalano PM, Linder B, Brickman WJ, Clayton P, Deerochanawong C, Hamilton J, Josefson JL, Lashley M, Lawrence JM, Lebenthal Y, Ma R, Maresh M, McCance D, Tam WH, Sacks DA, Dyer AR, Metzger BE; HAPO Follow‐up Study Cooperative Research GroupJAMA 2018;320: 1005–1016Glucotypes reveal new patterns of glucose dysregulationHall H, Perelman D, Breschi A, Limcaoco P, Kellogg R, McLaughlin T, Snyder MPLoS Biol 2018;16: e2005143Impact of daily incremental change in environmental temperature on beta cell function and the risk of gestational diabetes in pregnant womenRetnakaran R, Ye C, Kramer CK, Hanley AJ, Connelly PW, Sermer M, Zinman BDiabetologia 2018;61: 2633–2642Acceleration of BMI in early childhood and risk of sustained obesityGeserick M, Vogel M, Gausche R, Lipek T, Spielau U, Keller E, Pfäffle R, Kiess W, Körner AN Engl J Med 2018;379: 1303–1312Continuous glucose monitoring in pregnant women with type 1 diabetes: an observational cohort study of 186 pregnanciesKristensen K1,2,3, Ögge LE4,5, Sengpiel V4,5, Kjölhede K4,5, Dotevall A5,6, Elfvin A5,7, Knop FK8,9, Wiberg N1,3, Katsarou A10,11, Shaat N10,11, Kristensen L2, Berntorp K10,111Department of Clinical Sciences Lund, Lund University, Lund, Sweden; 2The Parker Institute, Copenhagen University Hospital, Copenhagen, Denmark; 3Department of Obstetrics and Gynecology, Skåne University Hospital, Malmö, Sweden; 4Department of Obstetrics and Gynecology, Sahlgrenska University Hospital, Gothenburg, Sweden; 5Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden; 6Department of Medicine, Östra/Sahlgrenska University Hospital, Gothenburg, Sweden; 7Department of Pediatrics, Sahlgrenska University Hospital, Gothenburg, Sweden; 8Clinical Metabolic Physiology, Steno Diabetes Center Copenhagen, Gentofte Hospital, Hellerup, Denmark; 9Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark; 10Department of Endocrinology, Skåne University Hospital, Malmö, Sweden; 11Department of Clinical Sciences Malmö, Lund University, Lund, SwedenDiabetologia 2019;62: 1143–1153BackgroundData relating continuous glucose monitoring (CGM) in type 1 diabetes pregnancy to neonatal outcomes are scarce. This real-world analysis of glycemic profiles from pregnant women using either real-time continuous glucose monitoring (rtCGM or intermittent CGM [iCGM]) aimed to better understand the associations between maternal CGM profiles throughout pregnancy and neonatal complications.MethodsData from 92 rtCGM and 94 iCGM pregnant users, were combined to provide detailed glycemic profiles in 186 Swedish women from two tertiary hospitals. Women made their own choice of which CGM device to use. 38% of iCGM users and 72% of rtCGM users were new to CGM during pregnancy. Maternal glycemic control was assessed by HbA1c in each trimester and CGM measures throughout pregnancy. The main neonatal outcomes were large for gestational age (LGA, defined as >2 SD above the expected birthweight), and at least one neonatal complication (macrosomia, shoulder dystocia, neonatal hypoglycemia, or admission to NICU >24 hours).ResultsOf 186 women in the study, 155 (83%) had CGM glycemic profiles in the first trimester, 165 (89%) in the second trimester, and 167 (90%) in the third trimester. The proportion of women with target HbA1c 7.8 mmol/L or 140mg/dL). Real-time CGM users spent significantly less time below range (3.5–7.8 mmol/L or 63–140 mg/dL) throughout pregnancy, with iCGM users spending strikingly high time below range in the late first and third trimesters. Overall, 52% and 53% of infants were born LGA with significant associations between mean second and third trimester CGM glucose and LGA (odds ratio [OR] 1.53 and 1.57, respectively). Higher second and third trimester TIR was associated with lower risk of LGA (OR 0.96 and 0.97). Higher second trimester glycemic variability (glucose SD) was associated with LGA and higher third trimester SD with neonatal complications.ConclusionsDespite 70% of women achieving target HbA1c levels, CGM data confirmed suboptimal day-to-day glucose control throughout pregnancy. A 5%–7% higher TIR during the second and third trimesters, was associated with decreased risk of LGA and neonatal complications, including macrosomia, shoulder dystocia, neonatal hypoglycemia, or NICU admissions of >24 hours in duration.CommentObjectively measured glucose control was suboptimal in this real-world cohort of iCGM and rtCGM users. A TIR target of 70% was achieved only in the final weeks of the third trimester, too late for optimal neonatal outcomes. Pregnant women with type 1 diabetes should be encouraged to aim for a TIR target of 70% and a TAR target of <25% as early as possible during the second and third trimester. For those who cannot reach 70%, small 5%–7% differences are associated with improved neonatal outcomes. More work is needed to support women and healthcare professionals to optimize use of technology before and during pregnancy.Maternal glycaemic control and risk of neonatal hypoglycaemia in type 1 diabetes pregnancy: a secondary analysis of the CONCEPTT trialYamamoto JM1,2, Corcoy R3,4, Donovan LE1,2, Stewart ZA5,6, Tomlinson G7, Beardsall K8,9, Feig DS10,11,12, Murphy HR5,13,14 on behalf of the CONCEPTT Collaborative Group1Departments of Medicine and Obstetrics and Gynaecology, University of Calgary, Calgary, Canada; 2Alberta Children's Hospital Research Institute, Calgary, Canada; 3Servei d'Endocrinologia i Nutrició, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 4CIBER‐BBN, Madrid, Spain; 5Wellcome Trust‐Medical Research Council Institute of Metabolic Science, University of Cambridge, Cambridge, UK; 6Department of Cardiovascular Sciences, University of Leicester, UK; 7Department of Medicine, University Health Network, Toronto, Canada; 8Department of Paediatrics, University of Cambridge, Cambridge, UK; 9Neonatal Unit, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK; 10Department of Medicine, University of Toronto, Toronto, Canada; 11Mount Sinai Hospital, Sinai Health System, Toronto, Canada; 12Lunenfeld‐Tanenbaum Research Institute, Toronto, Canada; 13Women's Health Academic Centre, Division of Women's and Children's Health, King's College London, London, UK; 14Norwich Medical School, University of East Anglia, Norwich, UKDiabet Med 2019;36: 1046–1053BackgroundContinuous glucose monitoring (CGM) provides a more detailed assessment of glycemic control than standard capillary glucose monitoring. This secondary analysis of CONCEPTT used CGM measures to examine the association between maternal glycemic control and neonatal hypoglycemia, a common complication of type 1 diabetes pregnancy.MethodsCONCEPTT was a multicenter randomized trial of CGM in women with type 1 diabetes who were pregnant or planning a pregnancy. Women from both the pregnant and planning pregnancy trials with a live birth were included in this cohort study (n=225). Glycemic control was assessed in the first, second, and third trimesters using HbA1c and CGM measures. Additionally, intrapartum glycemic control was assessed for women with available CGM data in the 24 hours prior to delivery. Measures of infant size, adiposity, and fetal hyperinsulinemia (as assessed by cord blood C-peptide concentrations) were also examined for associations with neonatal hypoglycemia (defined as a glucose <2.6 mmol/L and requiring intravenous dextrose).ResultsNeonatal hypoglycemia was diagnosed in 25% of infants in this cohort. Glycemic control as assessed by both HbA1c and TIR was associated with neonatal hypoglycemia in the second and third trimesters. There were no significant associations with CGM measures during the intrapartum period and neonatal hypoglycemia, though the number of women with intrapartum CGM data were small (n=33). Infant size (extreme large for gestational age), adiposity (skinfold thickness), and fetal hyperinsulinemia (cord blood C-peptide concentration) were significantly higher in neonates with hypoglycemia compared with those without.ConclusionThe association between both glycemic control the second half of pregnancy and measures of infant size and adiposity with neonatal hypoglycemia suggest that antepartum glycemic control plays an important part in the development of neonatal hypoglycemia. Additional studies are needed to determine the relative contribution of maternal antepartum versus intrapartum glycemic control to the risk of neonatal hypoglycemia.CommentNeonatal hypoglycemia remains a common complication following type 1 diabetes pregnancies. This study is consistent with other studies that demonstrate the importance of antenatal glycemic control in reducing the risk of neonatal hypoglycemia. Additionally, the association of infant size and adiposity with neonatal hypoglycemia further supports this. Using CGM, this study demonstrated that small improvements in TIR (5%–7%) are associated with less neonatal hypoglycemia. Healthcare providers and women with diabetes can focus on increasing CGM time in range during the latter half of pregnancy to reduce the risk of neonatal hypoglycemia.Pumps or multiple daily injections in pregnancy involving type 1 diabetes: a prespecified analysis of the CONCEPTT randomized trialFeig DS1,2,3, Corcoy R4,5, Donovan LE6, Murphy KE1,2,3, Barrett JFR7, Sanchez JJ7, Wysocki T8, Ruedy K9, Kollman C9, Tomlinson G3,10, Murphy HR11,12,13 on behalf of the CONCEPTT Collaborative Group1Mount Sinai Hospital, Sinai Health System, Toronto, Canada; 2Lunenfeld‐Tanenbaum Research Institute, Toronto, Canada; 3Department of Medicine, University of Toronto, Toronto, Canada; 4Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 5CIBER‐BBN, Zaragoza, Spain; 6University of Calgary, Calgary, Canada; 7Sunnybrook Research Institute, Toronto, Canada; 8Nemours Children's Health System, Jacksonville, FL; 9Jaeb Center for Health Research, Tampa, FL; 10University Health Network, Toronto, Canada; 11Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK; 12Department of Women and Children's Health, King's College London, London, UK; 13Department of Medicine, University of East Anglia, UKDiabetes Care 2018;41: 2471–2479BackgroundTight glycemic control is associated with improved pregnancy outcomes in women with type 1 diabetes. However, the effectiveness of insulin pump therapy versus multiple daily injections (MDI) in achieving tight glycemic control in pregnancy has not been clearly established.MethodsIn this prespecified secondary analysis of CONCEPTT, women using an insulin pump versus MDI were compared. The primary outcome was change in HbA1c from baseline to 34 weeks gestation. Additional outcomes included CGM measures, patient-reported outcome measures, and pregnancy complications.ResultsA total of 248 pregnant women (125 and 123 using a pump and MDI respectively) were included in this cohort study. Individuals using insulin pump therapy were more likely to be married or common-law (94% vs 81%; P=0.003), less likely to smoke (10% vs 21%; P=0.02) and more likely to be taking a preconception multivitamin (39% vs 25%; P=0.03). In the first trimester, there was no significant difference in HbA1c between individuals on pump or MDI (6.84±0.71% vs 6.95±0.58%; adjusted P=0.23). However, at 34 weeks individuals on pump therapy had a significantly larger decrease in HbA1c from randomization to 34 weeks than those on MDI (−0.32±0.65% vs −0.55±0.59%; adjusted P=0.001). Infants of pump users were more likely to be admitted to the neonatal intensive care unit (P=0.02) and slightly more likely to have neonatal hypoglycemia requiring intravenous dextrose (P=0.05).ConclusionsIn this cohort study of CONCEPTT participants, women on pump had significantly higher HbA1c in the second half of pregnancy and a smaller change in HbA1c from early to late pregnancy.CommentThis is one of the largest cohort studies comparing pump to MDI in pregnancy. Even with adjustment for confounders, cohort studies in pregnancy are subject to bias because of the differences between women who choose pump versus MDI. However, it is one of the few contemporary studies to suggest potentially less favorable neonatal outcomes between pump and MDI users. The increased neonatal hypoglycemia was likely related to a 5% lower TIR at 24 weeks. This contrasts with a previous Canadian cohort study, which found women using pump therapy had a lower HbA1c compared with those using MDI in the third trimester (6.5±0.5% vs 6.8±0.8%; P=0.002). However, women using pumps started pregnancy with lower HbA1c than those on MDI (6.9±0.7% vs 7.6±1.4%; P=0.001). Interestingly, insulin pumps users also had less of a change in HbA1c from early to late pregnancy (−0.4% vs −0.8%). In CONCEPTT, pump users had a 5% lower (TIR 63–140 mg/dL 53 vs 48%) at 24 weeks, despite comparable first and third trimester CGM glucose measures. The difference between pump and MDI users could be related to different eating patterns, and/or inadequate insulin dose adjustment in pump users during pregnancy. Regardless of the underlying cause, insulin pump users and their clinicians need to be more aggressive with insulin dose increments, as insulin resistance increases. Since only 25 women used pumps with low glucose suspend features in CONCEPTT, it remains to be seen if the newer pumps with partially automated features, or hybrid closed-loop systems will facilitate more aggressive insulin dose adjustment in the second and third trimesters.Diabetes technology use among pregnant and nonpregnant women with T1D in the T1D ExchangePolsky S1, Wu M2, Bode BW3, DuBose SN2, Goland RS4, Maahs DM5, Foster NC2, Peters AL6, Levy CJ7, Shah VN1, Beck RW21Barbara Davis Center for Diabetes, Aurora, CO; 2Jaeb Center for Health Research, Tampa, FL; 3Atlanta Diabetes Centers, Atlanta, GA; 4Naomi Berrie Diabetes Center at Columbia University Medical Center, New York, NY; 5Department of Pediatrics, Stanford University, Stanford, CA; 6Department of Internal Medicine, Keck School of Medicine of the University of Southern California, Los Angeles, CA; 7Department of Medicine, Division of Endocrinology, Diabetes, and Bone Disease, Icahn School of Medicine at Mount Sinai, New York, NYDiabetes Technol Ther 2018;20: 517–523BackgroundLittle is known about the glucose monitoring and insulin delivery systems that women with type 1 diabetes use to optimize glucose control before and during pregnancy. Data from the T1D Exchange clinic registry were combined with self-reported maternal and fetal health outcomes of women with a delivery in the previous year to examine contemporary diabetes technology use in a real-world setting.MethodsData collected from 4,340 female participants aged 18 to 45 years from 68 T1D Exchange centers during 2010–2013 were categorized as recently, previously, or never pregnant. Demographic data, details of diabetes technology use, and biomedical details were collected through self-reported questionnaires, chart reviews, and clinic medical records. Delivery and postpartum maternal infant outcomes among recently pregnant women were collected through self-reported questionnaires.ResultsThere were 214 (4.9%) recent pregnancies, 1,540 (35.5%) previous pregnancies, and 2,586 (59.6%) women who were never pregnant. Women were 28±9 years old, 84% white non-Hispanic, median (inter quartile range [IQR]) diabetes duration 13 years (8–20), median (IQR) HbA1c 7.9% (7.0–9.0%), and median (IQR) total daily insulin dose 0.6 units/kg (0.5–0.8). Women who were recently or previously were older (29.2 and 32.8 years) and had longer duration of diabetes (15 and 16 years) than women who were never pregnant (24.3 years old and 12 years duration). They also had higher rates of diabetes technology use (74%, 60%, and 58% for continuous subcutaneous insulin infusion) and (36%, 17%, and 12% for CGM) for recently, previously, and never pregnant women, respectively. Recently pregnant women had lower HbA1c levels (6.5%, 7.8%, and 8.0%) for recently, previously, and never pregnant. Rates of fetal loss were particularly high among recently and previously pregnant women (46% and 34% reported miscarriages; 6% and 3% reported stillbirths). Data from 130 recently pregnant women confirmed high rates of neonatal morbidity among liveborn infants (65% LGA, 43% neonatal intensive care unit admission, and 27% delivered at 140 mg/dL). It also confirms that small differences in maternal overnight glucose levels are associated with neonatal outcomes and adds to the growing literature of the objective CGM measures in the development of pregnancy complications. It explores the use of functional data analysis to analyze time-series CGM data at a population level in order to maximize the temporal information obtained. These techniques may be particularly applicable for women with GDM and those with less severe glycemic conditions for whom conventional CGM summary statistics and time-in-range measures may be insufficiently sensitive.Association of gestational diabetes with maternal disorders of glucose metabolism and childhood adiposityLowe WL Jr1, Scholtens DM1, Lowe LP1, Kuang A1, Nodzenski M1, Talbot O1, Catalano PM2, Linder B3, Brickman WJ1,4, Clayton P5, Deerochanawong C6, Hamilton J7, Josefson JL1,4, Lashley M8, Lawrence JM9, Lebenthal Y10, Ma R11, Maresh M12, McCance D13, Tam WH11, Sacks DA9, Dyer AR1, Metzger BE1, HAPO Follow‐up Study Cooperative Research Group1Northwestern University Feinberg School of Medicine, Chicago, IL; 2MetroHealth Medical Center, Case Western Reserve University, Cleveland, OH; 3National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD; 4Ann and Robert H. Lurie Children's Hospital, Chicago, IL; 5Royal Manchester Children's Hospital, Manchester University Hospitals NHS Foundation Trust, Manchester Academic Health Sciences Centre, School of Medical Sciences, Faculty of Biology, Medicine, and Health, University of Manchester, Manchester, UK; 6Rajavithi Hospital, Bangkok, Thailand; 7Hospital for Sick Children, University of Toronto, Toronto, Canada; 8Queen Elizabeth Hospital, School of Clinical Medicine and Research, University of the West Indies, Barbados; 9Kaiser Permanente of Southern California, CA; 10Jesse Z. and Sara Lea Shafer Institute for Endocrinology and Diabetes, National Center for Childhood Diabetes, Schneider Children's Medical Center of Israel and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel; 11Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong, China; 12St Mary's Hospital, Manchester University Hospitals NH
Referência(s)