Original Article Audit of antenatal steroid use in mothers of preterms admitted to a neonatal inten- sive care unit in Ghana. Adziri H. Sackey1 and Lily G. Tagoe2 Ghana Med J 2018; 52(2): 3-7 DOI: http://dx.doi.org/10.4314/gmj.v52i1.2 1Department of Child Health, School of Medicine and Dentistry, College of Health Sciences, University of Ghana, Accra, Ghana 2Department of Child Health, Korle Bu Teaching Hospital, Accra, Ghana. Corresponding author: Dr Adziri H. Sackey E-mail: sackey@sky.com Conflict of interest: None declared SUMMARY Background: Antenatal corticosteroids (ACS) are established as an effective method of reducing preterm morbidity and mortality. At the Korle Bu Teaching Hospital (KBTH), a tertiary referral centre in Ghana, it is recommended that a course of ACS should be given to mothers before delivery between 24 weeks to 34 weeks gestation. Objectives: The study was performed primarily to determine the level of adherence to guidelines on administration of ACS. Methods: All babies with gestational ages up to 34 weeks admitted to the neonatal unit (NICU) during the period of the study were eligible for inclusion. Results: There were 214 eligible admissions during the study period, of which 93 babies (43%) were studied due to poor access to medical records. Dexamethasone was the only steroid used, and mothers of 65 (70%) of the 93 babies received at least one dose; 17 (18%) received only one dose, 35 (38%) - 2 doses, 9 (10%) – 3 doses, and 4 (4%) - 4 doses. Conclusions: This study has shown a low uptake of antenatal corticosteroids, similar to other low-income and mid- dle-income countries. To improve preterm survival and morbidity, there is an urgent need to increase the use of cor- ticosteroids before preterm deliveries in Ghana and other low-income and middle-income countries. Funding: Not indicated Keywords: Antenatal corticosteroids, prematurity, neonatal intensive care, audit, low-income countries. INTRODUCTION Prematurity is recognised as a leading cause of neonatal in RDS is seen in infants born up to 7 days after the first morbidity and mortality.1 Liggins and Howie showed in dose; no reduction in neonatal death, RDS or cerebro- 1972 that administration of a single course of antenatal ventricular haemorrhage is seen in infants delivered corticosteroid to women at risk of preterm delivery re- more than 7 days after treatment with antenatal cortico- sulted in a lower incidence and severity of respiratory steroids.5 distress syndrome, intraventricular haemorrhage and necrotising enterocolitis.2 Antenatal corticosteroids This study was performed to determine the level of ad- (ACS) are now established as an effective and relatively herence and the impediments to compliance with guide- low-cost method of reducing preterm morbidity and lines on administration of ACS. It is hoped that this mortality.3 It is therefore important to monitor uptake of knowledge will help improve adherence to the guide- ACS in preterm delivery. lines. Furthermore, to help disseminate improved stand- ards of antenatal and neonatal care, the literature on use Antenatal steroids increase foetal surfactant production of antenatal corticosteroids in low-income and middle- through accelerated development of type 2 pneumo- income countries has been reviewed. cytes, and enhance the neonatal response to postnatal surfactant treatment.4 Their use appears to reduce neo- METHODS natal mortality rate even when infants are born less than The study was based on the neonatal unit (NICU) at the 24 hours after the first dose has been given. Reduction Korle Bu Teaching Hospital and performed in 2 phases: 3 www.ghanamedj.org Volume 52 Number 1 March 2018 Original Article from January to March 2015 and from December 2015 4. All patients who do not receive a full course to January 2016. The interruption in the study was due should have acceptable documented reasons. to inconsistent availability of staff for prompt identifica- tion of newly admitted babies and completion of ques- RESULTS tionnaires. When babies were recruited some days after The study involved preterm babies admitted to the admission, instead of immediately after admission, their KBTH NICU with gestations up to 34 completed weeks, mothers were more likely to have been discharged, and from January to March 2015 and from December 2105 their case notes were then impossible to retrieve. This to Jan 2016. During these periods there were 214 eligi- resulted in a significant number of otherwise eligible ble admissions, of which only 43% (93 babies) could be babies not being included in the audit. studied due to missing case notes. In the first phase, 52 out of 165 eligible babies (32%) were recruited, and in The details of all babies admitted to the NICU are rou- the second phase, 41 out of eligible 49 babies (84%). tinely entered into a register which, for the purposes of this audit, was inspected on a daily basis. All babies There were 51 males (55%) and 42(45%) females. Fif- with gestational ages of at most 34 weeks (by dates or ty-two babies (56% of the total) were delivered vaginal- examination) who had been admitted to NICU during ly. Maturity by dates is shown in table 1. Ten percent of the period of the study were eligible for inclusion in the the infants were born before 28 weeks gestation, and study. It is standard practice in the NICU for all admit- 87% between 28 and 34 weeks. ted babies who were preterm by dates to be assessed by experienced doctors for gestational age using the Bal- Table 1 Maturity by dates (MBD) of infants in the study lard scoring system.6 Collection of information was MBD Frequency Percentage standardised by the use of a questionnaire designed for 24 - 27 weeks + 6 days 9 10 the purpose. In the first phase, paediatric and obstetric 28 – 30 weeks + 6 days 33 35 residents attached to the NICU identified new admis- 31-34 weeks + 0 days 50 54 sions and completed these questionnaires by interview- Not entered *1 3 ing mothers and by reviewing clinical notes of babies. Total 93 100 They also attempted to retrieve and review case-notes of *The maturity by assessment of this baby was 31 weeks these mothers. In one case where data was unavailable from these sources, it was obtained verbally from moth- Birth weights are shown in table 2. Fifty eight percent er’s obstetrician. Due to difficulties with resident doc- of infants were less than 1500g, and 41% weighed be- tors’ availability to collect data, the study was suspend- tween 1500g and 2499g. ed until a dedicated research assistant was appointed to restart data collection in the second phase of the study. Table 2 Birth weights of the 93 infants in the study Birth weight Number Percentage Audit standards were based on guidelines in the De- <1000g 9 10 partment of Obstetrics and Gynaecology manual, Korle- 1000-1499g 45 48 Bu Teaching Hospital. 1500-1999g 29 31 2000-2499g 9 10 The relevant guideline statement was: 2500-3000g 1 1 A course of ACS should be given to all mothers before N 2 2 delivery between 24 weeks to 34 weeks gestation unless Total 93 100 the risk to the mother or foetus outweighs the benefit (chorioamnionitis, acute foetal distress, and acute mas- The commonest indications for preterm delivery were sive Antepartum haemorrhage). The course is IM Dex- premature rupture of membranes (22%), hypertensive amethasone 6mg 12hourly for 4 doses, or where availa- disorders managed by induction of labour or caesarean ble, Betamethasone 12mg IM, repeated 24 hours later. section (22%), and spontaneous preterm labour (11%). Seventy babies (75% of the total) were delivered in The standards for the study were KBTH, 20 (22%) in other facilities, and 3(3%) at home. 1. All mothers of preterm babies less than 34 Maternal ages are shown in table 3. Seven percent of the weeks gestation requiring steroids should have mothers were less than 21 years old and 4% older than received ACS. 40 years. 2. All ACS administered should be documented. 3. ACS should be given as IM dexamethasone 6mg 12hourly for 4 doses or IM betamethasone 12mg repeated 24 hours later. 4 www.ghanamedj.org Volume 52 Number 1 March 2018 Original Article Table 3 Maternal ages of infants in the study Rates of antenatal corticosteroid use varied between Age Frequency Percentage countries (median 54%, range 16–91%; IQR 30–68%).7 <18 years 3 3 Our rate was also higher than that obtained in a study 18-20 years 4 4 performed by Gwako G, 2012 in the Kenyatta National 21-25 years 11 12 Hospital in Kenya, where 46% of mothers who deliv- 26-30 years 33 35 ered prior to 34 weeks received antenatal corticoster- 31-35 years 15 16 oids.8 Other middle-income countries with low antenatal 36-40 years 17 18 steroid coverage, compared with this study, include >40 years 4 4 Cameroon (10%), Brazil (4%) and Ecuador (35%).9 Not document- 6 6 Pattanittum P et al reported that 40% of women who ed delivered at less than 34 weeks gestation in 9 hospitals Total 93 100 * in 4 countries in South East Asia, received antenatal * Rounded up from 98% corticosteroids.10 In Malawi, in a study by Ahlsen et al in two hospitals in the capital, Lilongwe – one district Eight (9% of 93 audited) mothers had a previous pre- and one tertiary, less than 10% of the babies studied term delivery. Sixty-five (70% of 93 audited) mothers received any antenatal corticosteroids.11 Comparing received antenatal steroids. Dexamethasone was the mothers who received a full course of steroids (4 doses steroid given in all 65 cases: 17(18% of 93) received 1 of dexamethasone), we obtained 4% in this study; dose, 35(38%) 2 doses, 9(10%) 3 doses, and 4(4%) re- Gwafo et al reported 3% in Kenya8; and Ahlsen et al 4% ceived 4 doses. in Malawi.11 Gwafo however inappropriately used the number exposed to antenatal steroids as the denomina- At least 74% of those who received antenatal steroids tor for this calculation instead of using the entire study had this documented in the mother’s clinical notes. It is population of 206, which would have resulted in a lower possible that there were more because in some cases figure of 1%. These results fall far short of our audit where antenatal corticosteroids were verbally reported standards and the practice in high-income countries. to have been given, the mothers’ notes were not availa- Among those who received antenatal corticosteroids, ble. Of those who received less than 4 doses, none had most had their last injection after 28 weeks gestation documented reasons as to why a full course of ANC was (92%). This was similar to that observed by Pattanittum not given. However, in 20 instances when steroids were P et al in 2008.10 not given, on detailed review of maternal notes and dis- cussion with obstetric colleagues – 6 arrived in second The commonest reasons for preterm delivery in our stage of labour, 5 required emergency caesarean section, study were spontaneous preterm labour, preterm prema- 1 was born just before arrival, 1 was a home delivery, ture rupture of membranes or hypertension among those and in 7 cases no reason was identified. Table 4 shows who received prenatal steroids. A similar finding was performance against the audit standards. obtained by Gwako G et al8 and Pattanittum et al. 10 Table 4 Assessing practice against standards Dexamethasone versus betamethasone Standard Target % compli- Two corticosteroids, dexamethasone and betame- (%) ance thasone, have been shown to be safe and effective in the At least one dose of ACS to 100 70 management of preterm birth.4 Roberts et al, after a eligible mothers Cochrane review of 21 studies, concluded that both cor- Full course of ACS given 100 4 ticosteroids significantly reduced combined foetal and Documentation of any ACS 100 74 neonatal death, RDS and cerebroventricular haemor- given rhage.5 Betamethasone however results in a greater re- Documentation of why full 100 0 duction in RDS while dexamethasone significantly in- course not given creased the incidence of puerperal sepsis. 5 These find- DISCUSSION ings were not in support of an earlier trial by Jobe et al which demonstrated a decreased risk of death with be- In this study, 70% of babies delivered between 24 and tamethasone but not dexamethasone.12 Lee et al also 34 weeks gestation received at least one dose of antena- corroborated the findings by Jobe.13 The World Health tal dexamethasone. This was higher than the 52% ob- Organisation (WHO) does not express a preference.14 tained in the WHO Multicountry Survey on Maternal Dexamethasone is the antenatal corticosteroid common- and Newborn Health, performed in 29 countries and ly available in Ghana; the price for a 4mg ampoule involving 7547 women who gave birth at 26-34 weeks’ ranges from 0.80 to 1.20 Ghana cedis. gestation. 7 5 www.ghanamedj.org Volume 52 Number 1 March 2018 Original Article Betamethasone is currently unavailable in the Korle-Bu Prediction of imminent delivery Teaching Hospital but was available in April 2016 at Accurate prediction of preterm delivery is essential to one private pharmacy at 73 Ghana cedis for a 2mg am- ensure timely administration of antenatal corticoster- poule; a course of Betamethasone (2 doses of 12mg) oids. In a study by Gyamfi-Bannerman et al in 2016, the therefore cost 876 Ghana cedis, compared to about 6 following features were considered indicative of a high Ghana cedis for Dexamethasone. It is likely that Beta- probability of delivery: spontaneous rupture of mem- methasone may be obtained at a lower price if pur- branes; preterm labour with at least 3cm dilatation or chased in bulk directly by the hospital. 75% cervical effacement; or any other indication for preterm delivery by induction or caesarean section.19 Benefit of ACS in absence of neonatal intensive care Transvaginal ultrasound assessment for cervical length Regarding the widespread use of ACS, cautions have and funnelling is another means of predicting imminent been raised by some studies. Azad et al, after a system- preterm birth. atic review of 21 studies in high-income and middle- income country hospitals, concluded that the observed Limitations of this study reductions in neonatal mortality rates after antenatal A limitation of this study is the relatively small study corticosteroids were highly unlikely to be replicated in population caused by 1) difficulty in retrieving maternal settings without level 2 neonatal care. They recom- clinical notes once the mothers were discharged 2) una- mended restriction of single-dose antenatal steroids to vailability of some mothers to complete questionnaires mothers at “33 weeks’ gestation or less, in preterm la- and 3) incomplete filling of some of the questionnaires bour, and with easy access to good quality, round-the by doctors. These difficulties were later overcome by clock level 2 care”.15 The American Academy of Paedi- the use of dedicated data collection personnel who ex- atrics stipulates that level 2 facilities should be capable tracted data before maternal discharge. This change re- of delivering continuous positive airway pressure and sulted in a higher recruitment of eligible cases - 52 out providing mechanical ventilation for up to 24 hours. of 165 (31.5%) in the first phase, and 41 out of 49 They must also have equipment such as portable X-ray (84%) with the help of a research assistant in the second machines and blood gas analysers, as well as various phase. Babies who received antenatal steroids but were personnel such as physicians, specialised nurses, respir- delivered after 34 weeks gestation were not included in atory therapists, radiology and laboratory technicians, this study. Also excluded were stillbirths and preterm who must be available at all times to manage emergen- infants who did not survive long enough to be admitted cies.16 The KBTH NICU is not quite up to level 2 but to NICU. has level 3 responsibilities, such as the care of babies born at <32 weeks’ gestation, as well as those weighing Conclusion and Recommendations less than 1500g. Azad’s concerns seem to have been The use of antenatal corticosteroids in KBTH compares borne out by the findings of Althabe et al (2015). They favourably with data from other low-income and mid- reported, in a trial in low and middle-income countries, dle-income countries but is far below acceptable levels. that increased use of antenatal corticosteroids in low The limited data from referring facilities indicates even birth weight infants did not decrease neonatal mortality, lower usage of antenatal corticosteroid, and this must be and that for every 1000 women exposed to this strategy, addressed by increasing awareness of health workers of an excess of 3.5 neonatal deaths occurred, and maternal the indications for the use of ACS. This task is rendered infection seemed to increase.17 However, a significant more difficult by inaccurate dating of pregnancy due to weakness of this study was the use of low birth weight lack of formal education and poor access to antenatal as a proxy for prematurity, which could result in inap- care. Improving the uptake of ACS is daunting enough propriate steroid exposure for many pregnancies. but it is only part of the package of measures needed to achieve better neonatal outcomes. For example, the full Multiple courses of ACS benefits of ACS may not be realised if the standards of According to current American College of Obstetricians neonatal intensive care are low. Training should be of- and Gynaecology guidelines (2011), a single rescue fered to health workers on prediction of preterm deliv- course of antenatal corticosteroids may be considered if ery to help reduce instances where delivery does not the gestational age is less than 32 weeks+6 days and the occur after administration of steroids resulting in the initial course of antenatal steroids was given more than need to consider multiple courses of ACS. 2 weeks prior and the woman is judged to be likely to give birth within the next week. Regularly scheduled To facilitate implementation of the WHO recommenda- repeat courses or more than two courses are not recom- tion “that dexamethasone or betamethasone should be mended.18 available in all maternity facilities and should be includ- ed in national essential drugs lists“13, we suggest that 6 www.ghanamedj.org Volume 52 Number 1 March 2018 Original Article routine documentation of steroid use should be made yatta National Hospital, J Obstet Gynaecol East mandatory in order to facilitate regular audit; health Cent Africa. 2013; 25(1): 3-9 facilities should strengthen their medical records de- 9. Liu G, Segrè J, Gülmezoglu AM, Mathai M, Smith partments and employ research assistants to enable col- JM, Hermida J et al. Antenatal corticosteroids for lection of high quality data for monitoring of perfor- management of preterm birth: a multi-country anal- mance. ysis of health system bottlenecks and potential solu- tions, BMC Pregnancy Childbirth. 2015; 15(Suppl ACKNOWLEDGEMENT 2): S3 Dr Ozoya, Dr Mustapha and Dr Insaidoo helped with 10. Pattanittum P, Ewens MR, Laopaiboon M, Pisake initial questionnaire design and data analysis. The fol- Lumbiganon, McDonald SJ, Crowther CA, Use of lowing paediatricians and obstetricians undertook initial antenatal corticosteroids prior to preterm birth in data collection: Dr Okaikwei Amartey, Dr L. Batsa- four South East Asian countries within the SEA- Nakotey, Dr N. Okai Brako, Dr Agbley Harry Benson, ORCHID project, BMC Pregnancy Childbirth. Dr Birikorang, Dr Mensah and Dr Mintah-Yenzu. Dr 2008; 8: 47 Samuel Oppong, consultant obstetrician KBTH provid- 11. Ahlsen AK, Spong E, Nomsa K, Kamwendo F, ed helpful comments on the final manuscript. Wolff K, Born too small: who survives in the pub- lic hospitals in Lilongwe, Malawi?, Arch Dis Child REFERENCES Fetal Neonatal Ed. doi:10.1136/archdischild-2013- 1. Bhutta ZA and Black RE. Global Maternal, New- 305877 born, and Child Health — So Near and Yet So Far. 12. Jobe AH, Soll RF. Choice and dose of corticoster- N Engl J Med. 2013; 369:2226-35 o i d f o r a n t e n a t a l t r e a t m e n t s . A m J O b s tet Gynecol. n engl j med 369;23 nejm.org december 5, 2013 2004; 190:878. 2. Liggins GC, Howie RN. A controlled trial of ante- 13. Lee BH, Stoll BJ, McDonald SA, Higgins RD. Ad- partum glucocorticoid treatment for prevention of verse neonatal outcomes associated with antenatal the respiratory distress syndrome in premature in- dexamethasone versus antenatal betamethasone. fants. Pediatrics. 1972; 50(4):515-25. Pediatrics. 2006; 117:1503 3. Mwansa-Kambafwile J, Cousens S, Hansen T, 14. Hofmeyr GJ. Antenatal corticosteroids for women Lawn JE. Antenatal steroids in preterm labour for at risk of preterm birth: RHL commentary (last re- the prevention of neonatal deaths due to complica- vised: 2 February 2009). The WHO Reproductive tions of preterm birth. Int J Epidemiol. 2010 Apr; Health Library; Geneva: World Health Organiza- 39(Suppl 1): i122–i133. tion. 4. Lee M-J, Guinn D, Antenatal corticosteroid therapy 15. Azad K, Costello A. Extreme caution is needed for reduction of neonatal morbidity and mortality before scale-up of antenatal corticosteroids to re- from preterm delivery, UpToDate. 2015, duce preterm deaths in low-income settings. Lancet http://www.uptodate.com/contents/antenatal- Glob Health. 2014;(2):e191-192 corticosteroid-therapy-f…reduction-of-neonatal- 16. Committee on Fetus and Newborn, Levels of Neo- morbidity-and-mortality-from-preterm-delivery natal Care, Paediatrics. 2012; 130:587-597 (Accessed on 16/FEB/2016) 17. Althabe F, Belizán JM, McClure EM, Hemingway- 5. Roberts D, Dalziel SR. Antenatal corticosteroids Foday J, Berrueta M, Mazzoni A et al, A popula- for accelerating fetal lung maturation for women at tion-based, multifaceted strategy to implement an- risk of preterm birth. Cochrane Database Syst Rev tenatal corticosteroid treatment versus standard care 2006;3:CD004454. for the reduction of neonatal mortality due to pre- 6. Ballard JL, Khoury JC, Wedig K, Wang L, Eilers- term birth in low-income and middle-income coun- Walsman BL, Lipp R. New Ballard Score, expand- tries: the ACT cluster-randomised trial, The Lancet. ed to include extremely premature infants. J Pedi- 2015, 385(9968): 629-639 atrics. 1991 Sep; 119(3): 417-423. 18. American College of Obstetricians and Gynaecol- 7. Vogel JP, Souza JP, Gülmezoglu AM, Mori R, ogists (ACOG) Committee on Obstetric Practice, Lumbiganon P, Qureshi Z et al. Use of antenatal Obstet Gynecol. 2011 Feb; 117(2 Pt 1): 422-424 corticosteroids and tocolytic drugs in preterm births 19. Gyamfi-Bannerman C, Thom EA, Blackwell SC, in 29 countries: an analysis of the WHO Multicoun- Tita ATN, Reddy UM, Saade GR et al, Antenatal try Survey on Maternal and Newborn Health. The Betamethasone for women at risk for late preterm Lancet. 2014; 384(9957): 1869-1877 delivery N Engl J Med. 2016 Feb, DOI: 8. Gwako G, Qureshi ZN, Kudoyi W, Were F. Ante- 10.1056/NEJMoa1516783 (Accessed on natal corticosteroid use in preterm delivery at Ken- 10/FEB/2016) ✪ 7 www.ghanamedj.org Volume 52 Number 1 March 2018