Newborn screening for sickle cell disease: an innovative pilot program to improve child survival in Dar es Salaam, Tanzania

Abstract

Background

Sickle cell disease (SCD) is a recognized cause of childhood mortality. Tanzania has the fifth highest incidence of SCD (with an estimated 11 000 SCD annual births) worldwide. Although newborn screening (NBS) for SCD and comprehensive healthcare have been shown to reduce under-5 mortality by up to 94% in high-income countries such as the USA, no country in Africa has maintained NBS for SCD as a national health program. The aims of this program were to establish and evaluate NBS-SCD as a health intervention in Tanzania and to determine the birth prevalence of SCD.

Methods

Muhimbili University of Health and Allied Sciences conducted NBS for SCD from January 2015 to November 2016. Dried blood spot samples were collected and tested for SCD using isoelectric focusing.

Results

Screening was conducted on 3981 newborns. Thirty-one (0.8%) babies had SCD, 505 (12.6%) had sickle cell trait and 26 (0.7%) had other hemoglobinopathies. Twenty-eight (90.3%) of the 31 newborns with SCD were enrolled for comprehensive healthcare.

Conclusions

This is the first report on NBS as a health program for SCD in Tanzania. The SCD birth prevalence of 8 per 1000 births is of public health significance. It is therefore important to conduct NBS for SCD with enrollment into a comprehensive care program.

comprehensive care, newborn screening, sickle cell disease, Tanzania

Introduction

Sickle cell disease (SCD) refers to a group of inherited red blood cell disorders that occur due to mutations in beta globin genes, one of which is invariably hemoglobin S (HbS), a variant produced as a result of a glutamic acid to valine substitution at position 6 of the beta globin chain, the underlying mutation being HBB:c.20A>T. HbS is prone to polymerization, disrupting red blood cell shape, function and life span. Individuals who are heterozygous for HbS (sickle cell trait) are usually asymptomatic while those who are homozygous have SCD. SCD can also occur when HbS is coinherited with beta thalassemia and other hemoglobin variants such as HbC. However, homozygous HbS and coinheritance with beta zero thalassemia result in the most severe forms. SCD is characterized by chronic hemolytic anemia and recurrent vaso-occlusions, which lead to painful crises, the hallmark of the disease.

Worldwide it is estimated that >300 000 babies are born annually with SCD and that these numbers will increase from 305 800 in 2010 to 404 200 in 2050.¹ Most of these babies are born in sub-Saharan Africa and India^2–4 where SCD contributes significantly to early childhood mortality and morbidity.^5,6 In 2006, the WHO, at its 59th health assembly, identified SCD as a significant public health burden in Africa.^7,8 Tanzania is estimated to rank fifth after Nigeria, the Democratic Republic of Congo, India and Angola as countries with the highest SCD birth prevalence.³ Up to 11 000 SCD births occur annually in Tanzania and, without proper diagnosis and management, 50% or more of these babies will die before the age of 18 y, with the highest mortality in the under-5 (U5) age group.⁶ The high SCD prevalence is also reflected in the high proportion of individuals who are carriers of the sickle cell gene, also known as sickle cell trait (SCT) (13–20%),⁶ in areas including the eastern coast and the northwest region around Lake Victoria.⁹ In recognition of this condition, the country has included SCD as a disease of public health priority in the National Noncommunicable Disease Strategy.¹⁰

The mortality and morbidity associated with SCD can be reduced by early preventive measures including newborn screening (NBS) coupled with comprehensive care and educative programs for affected families. There is significant evidence from high-income countries of increased survival for individuals with SCD following early diagnosis by NBS and comprehensive care.^11–14 In Africa, although there have been different initiatives for NBS-SCD in Ghana, Nigeria, Angola and Uganda,^15–18 no country has successfully established NBS-SCD as a universal national health intervention. Ghana has the largest and the most extensive initiatives. In 2015, Tanzania, through the Muhimbili University of Health and Allied Sciences (MUHAS), implemented a pilot NBS for SCD program in Dar es Salaam, Tanzania. The aims of this pilot program were to establish and evaluate NBS-SCD as a health intervention in Tanzania and to determine the birth prevalence of SCD. The innovation of this program was based on utilizing expertise and resources from existing public health programs in Tanzania such as Management and Development for Health (MDH), Comprehensive Community Based Rehabilitation in Tanzania (CCBRT) and the Ifakara Health Institute (IHI).

This pilot program adopted the NBS for SCD model from Public Health England (PHE) and customized it to fit our settings. Key issues included location of the laboratory, dried blood spot (DBS) collection and transportation, laboratory screening tests, feedback of results and turnaround times. Globally, NBS-SCD programs have employed a variety of models. In most African settings, where there is a short interval between time of delivery and that of discharge, most NBS programs are designed to capture babies before discharge.¹⁶ However, in other settings, it has proven beneficial to allow flexible capturing times up to a few weeks after birth.¹⁵ The type of sample collected for screening is also an important factor. Although most NBS programs use DBS from a heel prick, some have also tried using umbilical cord samples.¹⁹ Sample transportation can be a challenge, especially in underdeveloped settings. However, initiatives that have been successful have included the use of postal offices and couriers. Screening test choice forms an important basis of NBS for SCD. Most African countries have employed isoelectric focusing (IEF) as the first-line test and another standard test such as high performance liquid chromatography (HPLC) or capillary electrophoresis (CE). Feedback of results in most settings has utilized mobile phone communication; however, this has been challenging when mothers could not provide correct numbers or did not have access to a mobile phone. The details of how our program was conducted are described below.

Materials and methods

Study area and population

Study area

This was a prospective study involving newborns at Muhimbili National Hospital (MNH) and Temeke Regional Hospital, Dar es Salaam, Tanzania, from January 2015 to November 2016. MNH is the nation’s tertiary health facility and has been offering SCD services, both clinical and laboratory, for the past three decades, while Temeke Regional Hospital, which is a regional hospital for the Temeke municipality in Dar es Salaam, established SCD services more recently. Both MNH and Temeke Regional Hospital each provide services for residents in Dar es Salaam and the neighboring regions.

Study population

This study involved women who delivered at MNH and Temeke Regional Hospital, and screening involved newborns aged up to 2 d. Informed consent was requested from any woman who agreed for her newborn to be screened. The study excluded all newborns with any illnesses and those whose mothers did not agree to participate in the study. Demographic and contact details were collected prior to the collection of the blood sample using a standard proforma and these were later entered into a project database developed in MySQL (Sun Microsystems Inc., Santa Clara, CA, USA).

Health education and training

Health education and training consisted of the provision of health education to mothers and the training of healthcare workers.

Education to mothers

Trained nurses provided health education to mothers who had delivered at MNH and Temeke Regional Hospital in the postnatal and neonatal wards. Health education was delivered before obtaining consent for a newborn to be screened. The duration of the education ranged from 30 min to 1 h depending on discussions/questions from the mothers. Health education on SCD was delivered through a brief talk with a group of mothers and by the provision of leaflets. The information included the origin of SCD and different types of sickle status, including the difference between a baby who is a carrier and one with SCD. Training was also provided for SCD health education including the importance of clinic visits and healthy living. The training was delivered in the national language of Swahili to ensure understanding. Training materials were based on existing resources developed for the Sickle Cell Program (SCP) of Tanzania.

Training of healthcare workers

Training on SCD care was provided to different cadres of health workers (doctors and nurses) identified at MNH and regional hospitals including Temeke, Ilala and Amana. Areas of strengthening included origin and inheritance of SCD, SCD screening (including the collection of DBS, specifically for nurses), diagnosis, neonatal health and SCD management. Training was scheduled for 5 d and training materials included brief talks, leaflets and videos, which were adopted from existing SCD resources developed by the SCP program in Tanzania and by PHE.

Training of laboratory technicians

Laboratory training materials and standard operating procedures were developed and used to train newly recruited NBS program laboratory technicians as well as laboratory technicians from MUHAS and Bugando Referral Hospital. The initial training by our UK partners (PHE) was conducted over 1 wk and covered the theoretical aspects of laboratory organization and management, sample collection, screening methods and unterpretation of results. Following the installation of IEF and later on HPLC, 2 d of training were provided and facilitated by the manufacturers, PerkinElmer Massachusetts, USA and Biorad California, USA, training the laboratory technicians on IEF and HPLC methods, respectively. In addition, the technicians underwent practical training on the use of the DBS puncher–user interface.

NBS for SCD

Enrollment into the screening program

Pregnant women who received SCD health education were requested to provide consent for their babies to be screened for SCD. Enrollment was completed using the screening registration form together with detailed information on how to locate the mother after the screening. The information collected included the location, street name, house number and phone numbers of spouse, parents and responsible person. All babies were given unique identification numbers.

DBS collection

The DBS sample was collected from the newborn’s heel using standard prickers. DBS collection, transportation and storage were performed according to the manufacturer’s protocols (Lasec Diagnostics), Ndabeni, South Africa and NBS guidelines from the Clinical and Laboratory Standards Institute.

SCD screening

Screening for SCD was performed at MUHAS hematology clinical research laboratory: 3.2 mm of the DBS was punched by an automated DBS puncher (Wallac DBS Puncher) PerkinElmer, Massachusetts, USA followed by IEF analysis, conducted according to the manufacturer’s protocol (Wallac RESOLVE Hemoglobin System). Samples that had abnormal results such as those found with the sickle gene (FAS), those found to be positive for SCD (FS) and those found to have unidentified variants FAV for the first test were repunched and retested by IEF. Results were interpreted initially by a trained laboratory technician and confirmed by the laboratory manager. Confirmed results were entered into the program’s database using the unique identification number.

Dissemination of results

Mothers or nominated close relatives of the screened babies who had a negative screening test for SCD received a standard text message informing them of the results, while those carrying a sickle gene (SCT) or those with SCD were contacted by a telephone call. Telephone calls were conducted following standard operating procedures, which included that (if necessary) the information should be passed to the responsible person and also a plan to follow in the case of a failed call or an unreachable number.

Comprehensive SCD healthcare

Clinical services for SCD had been previously established at MNH and Temeke Regional Hospital. SCD services were strengthened following training of the healthcare workers on SCD diagnosis and management. In addition, the program provided guidelines for SCD management at outpatient clinics and during hospitalization, with criteria for referral to MNH. Upon diagnosis of SCD, newborns were enrolled for a comprehensive healthcare assessment in these two hospitals. A simple case report form was established for each child with SCD when attending the SCD clinics. Reminders about the scheduled visits were sent by text messages to a primary and to an alternative number provided during clinic visits. Active surveillance was established as a means to track SCD children who had not attended a clinic after 6 months had passed.

Results

Health education and training

Health education for pregnant women/mothers of newborns

About 4000 pregnant women/mothers who had delivered at MNH and Temeke Regional Hospital received health education on SCD.

Training of healthcare workers

A total of 160 nurses and 96 healthcare workers from MNH, Temeke Regional Hospital, Amana and Mwananyamala hospitals were trained on the origin and inheritance of SCD, SCD screening/diagnosis, neonatal health and SCD management. With the establishment of NBS into national policy, the training materials will be incorporated into the syllabus of midwifery training so that it is provided as part of preservice training in nursing courses.

Training of laboratory technicians

A total of 16 laboratory technicians from Bugando Hospital, MNH and MUHAS laboratories were trained to perform SCD screening using IEF and HPLC methods. The laboratory technicians were selected from hospitals with either IEF or HPLC testing platforms. The training materials were submitted to the Ministry of Health, Community Development, Gender, Elderly and Children MoHCDGEC) for potential revision and approval for use.

NBS for SCD

SCD screening

DBS collection was conducted for 11 months (October 2015 to September 2016) following the establishment of the program. As this was a pilot phase and NBS is not a health policy in Tanzania, DBS collection was conducted based on the availability of resources and for a few hours each day. MNH and Temeke Regional Hospital each have about 800 and 1000 births per month, respectively. Overall, a total of 3981 babies were screened (1141 from MNH and 2840 from Temeke Regional Hospital). Although screening took place in two health facilities, the catchment area included all parts of Dar es Salaam as well as peripheral areas (Table 1), 68.9% coming from the Temeke municipality. Of the 3981 babies screened, 3419 (85.9%) were found to be FA, 505 (12.7%) FAS, 31 (0.8%) were found to be SCD-positive and 26 (0.7%) had unidentified variants. The prevalence of SCD positivity was similar between the MNH (0.61%) and Temeke Regional Hospital (0.85%; p=0.452). There was an equal proportion of males (50.7%) and females (49.3%) among the babies who were screened. However, out of the 31 babies who were found to be positive, the prevalence in females was 1.15% (22 out of 1910) while that of males was 0.41% (8 out of 1967); 104 babies (one SCD-positive) had their sex status missing.

Table 1.

Results of newborn screening by regions within and outside Dar es Salaam

	Ilala	Kinondoni	Temeke	Others*	Total number of babies
FA	369 (86.21%)	486 (89.5%)	2355 (85.89%)	209 (77.99%)	3419
FAS	52 (12.15%)	52 (9.58%)	347 (12.65%)	54 (20.15%)	505
FS	2 (0.47%)	3 (0.55%)	24 (0.88%)	2 (0.75%)	31
Others	5 (1.17%)	2 (0.37%)	16 (0.58%)	3 (1.12%)	26
Total	428 (100%)	543 (100%)	2742 (100%)	268 (100%)	3981

FA: normal babies; FAS: babies with sickle cell trait (carrier); FS: babies with SCD; Others: babies with unidentified variants; *: regions other than Dar es Salaam.

Feedback of test results and enrollment of SCD cases into comprehensive care

Of the 3981 screened babies, >80% of parents/guardians could be contacted and informed of the results; this included parents of all babies found to be carriers and 30/31 of the parents of babies with SCD. The NBS program followed up the 31 babies who were diagnosed with SCD at birth and ensured that they received comprehensive care for SCD. Of the 31 newborns diagnosed with SCD, 28 were enrolled into comprehensive care, two died and one could not be reached. The comprehensive care for SCD included the following: (1) SCD confirmation: laboratory tests for confirmation of SCD diagnosis and full blood count; (2) basic interventions: SCD patients received folic acid tablets (0.1–0.5 mg per day), prevention, prompt diagnosis and treatment of malaria and prevention of bacterial infection with daily oral penicillin V; and (3) health education: health information about SCD to parents or caregivers, advice to attend an outpatient clinic every 3 months and advice to seek medical care in the event of acute illness.

Discussion

Despite the evident benefits of NBS for the survival and well-being of individuals with SCD in developed countries, sub-Saharan countries with the greatest burden of disease are failing to establish similar national NBS programs. Several African countries including Ghana, Angola, Uganda and Nigeria have embarked on initiating NBS programs; however, none has been adopted as national screening programs. Among these countries, Ghana has achieved the largest screening initiative, with >500 000 babies screened up to 2017. There are a number of obstacles to achieving the objective of including NBS services for SCD as national programs, including challenges with health infrastructure, the disproportionately high cost of screening tests in the African setting and mechanisms for feedback of results.

This is the first report to document NBS establishment as a health program for SCD in Tanzania, which has the fifth highest incidence of SCD globally. The findings from this project have proven the importance of conducting NBS for SCD as a health intervention in countries with a high prevalence of SCD such as Tanzania. NBS services were initially established at MNH and thereafter Temeke Regional Hospital, with an ongoing plan to scale up to other regional hospitals. The successes and challenges that have been realized and documented in this report will be used to inform the Tanzanian government as strategies for a national NBS program are put in place.

Health education and training

Health education has been highlighted as an important contributor to the success of NBS-SCD initiatives.^15–17 We also experienced the importance of ongoing heath education and training for the families and healthcare workers. As part of the NBS program, health education was provided for mothers of newborns. In both MNH and Temeke Regional Hospital education was provided in postnatal and neonatal wards. This was a limitation to the number of women who benefited from the education and it is evident that the use of antenatal clinics may increase that number. At the moment, the health system in Tanzania allows for antenatal clinics in peripheral hospitals but not in referral hospitals such as MNH and Temeke Regional Hospital. Conducting NBS for SCD in peripheral hospitals such as Temeke Regional Hospital may therefore increase both the number of pregnant women who receive SCD health education and the number of babies screened. Another limitation was the mode that was used to deliver education, brief talks and leaflets. At the time when mothers have just given birth, it is difficult to concentrate on new information and to be able to remember this afterwards. Therefore, ongoing SCD education programs via media and at SCD clinics are important.

The program trained health workers from the selected hospitals and healthcare centers. The identification of healthcare workers was left to the hospital management to increase buy-in and sustainability. The healthcare workers included doctors and nurses from obstetrics, postnatal and neonatal wards. The number of pregnant women/newborns in the wards and the time spent in that ward before discharge dictated the selection of health workers for NBS. The modality of training the healthcare workers received followed the training of trainers approach, such that subsequent training sessions involved the trained personnel as trainers. Considering the low awareness of SCD among most healthcare workers in Tanzania, the focus of the training was general knowledge of SCD and disease management. Interactive talks and demonstrations featuring real settings were the best delivery methods rather than formal talks. The main challenge with the training of healthcare workers is the transfer of the trainees from the designated facilities to other facilities. Therefore, well-planned ongoing refresher training sessions for healthcare workers are required in the initial years of NBS until it is well established.

NBS for SCD

SCD screening

We observed that most (68.7%) of the babies that were screened reside in the Temeke district. The reason for this could be due to the high screening rate at Temeke Regional Hospital (71.3%), which is likely to serve the majority of the population living in the Temeke municipality. Furthermore, MNH being the national referral hospital is likely to receive at-risk mothers who are likely to be distributed evenly across the region. Although the prevalence was not significantly different between the two facilities (0.61% in MNH vs 0.85% in Temeke Regional Hospital), most likely due to limited sample size, this could also reflect differences in SCD prevalence across regions (SCD regional pockets) within Tanzania, which also is reflected in settlements in the capital city of Dar es Salaam. This is similarly true for the coastal (Dar es Salaam, Pwani, Tanga) and Lake Victoria (Mwanza and Mara) regions where SCD is highly prevalent.¹⁵ Although the proportion of babies screened was equal between boys and girls, the proportion of girls found with SCD was much higher (71%) than that of boys in the babies found to be positive for SCD. The overall proportion of SCD (0.8%) was more than the predicted value (0.6%), which may indicate an increase in SCD prevalence in this region or an underestimate nationally. In addition, the proportion of babies who are carriers of sickle cell (SCT) was high (12.5%), calling for preventive action for this group. This frequency is in line with previous reports.⁶ We also observed abnormal variants (0.7%) that could not be identified by the available screening methods, highlighting the need for DNA diagnostic tests that can resolve rarer hemoglobin variants.

Feedback of results to patients

Of the families of the screened babies, >80% could be contacted and given the results by phone; however, some of the parents could not be contacted, either due to a wrong telephone number or a misplaced/lost phone. This highlights the importance of investing in other methods such as use of a global positioning system or integration with the immunization registry, which has a well-established electronic system to capture and track immunization events.²⁰

Comprehensive care

Of the 31 identified babies, two had died, upon whom verbal autopsies were performed, and their respective probable causes of deaths indicated that the deaths were not associated with SCD but with other causes of deaths including hypothermia and hypoglycemia. The remainder of the babies were successfully enrolled for comprehensive care.

Development of strategy for NBS in Tanzania

Scaling up

The experience from this project will be used to strengthen or scale up of NBS in Tanzania. Conducting NBS first at MNH and then at Temeke Regional Hospital gave us an experience of scaling up and operating in both a national referral hospital and a regional hospital. The scaling-up model has been used elsewhere as this allows for proper establishment in one site prior to the involvement of multiple sites.²¹ The ultimate goal of the NBS program is to be a sustainable national health program owned by the government. The program has initiated discussions with the MoHCDGEC to ensure sustainability of NBS services. The approach that has proven to be sustainable for Tanzania is the integration of NBS services with existing healthcare systems in provision of health education, sample collection and dissemination of results. We have identified the potential best platforms to work with, one of which is the immunization program, and also by utilizing non-communicable diseases (NCD) clinics. The NCD strategic plan (2016–2020) (https://www.worlddiabetesfoundation.org/files/tanzania-ncd-stategic-plan-2016-2020) has the following priority areas: community sensitization and early detection by NBS, SCD screening at health facilities among newborns in high prevalence areas and provision of comprehensive care. Policy on the timing of SCD screening of pregnant women, newborns or as early infant diagnosis needs to be established.

National initiatives

At the national level, the SCP has worked with the government to establish a national SCD taskforce, which will work on implementing a national NBS program that will include SCD screening. The program also supported the MoHCDGEC in the development of the first draft of National Newborn Screening Guidelines, which includes SCD. The SCP continues to work with the ministry of health and other stakeholders to finalize the national NBS guidelines and to lobby for their implementation.

Limitations

The program faced challenges and limitations at a number of levels. The first was obtaining buy-in from program stakeholders and implementers. This was mitigated by conducting frequent meetings to share the development and progress of the program at site, regional and national levels. The second was delayed procurement and the high cost of laboratory equipment and reagents. Although not solved during the program, the mitigation for this lies in exploring different screening platforms, including point of care tests. Third, there were challenges in mechanisms of feedback of results due to poor tracking systems and dynamics of mobile phones. This has been addressed by exploring potential interoperability with the Tanzania electronic Immunization Registry.

Conclusions

Tanzania has made good progress on the targets for Millennium Development Goal 4, including a decline in U5 deaths from 166 in 1990 to 112 in 2005 and 67 in 2015 per 1000 live births. In addition, infant mortality decreased from 68 to 43 per 1000 live births between 2005 and 2015. NBS for SCD is expected to contribute to further progress in reducing infant mortality. The importance of NBS intervention in Tanzania is supported by previous data showing an increasing number of babies born with SCD in the country.¹ Despite the fact that NBS requires a significant investment, the cost-effectiveness of NBS across sub-Saharan African countries is based on the high prevalence of SCD.¹³

The main lesson is that an NBS program is needed and is practical in Tanzania. In the 11 months of piloting NBS at MNH and Temeke Regional Hospital (October 2015 to September 2016), there was success in counseling women at delivery, collection of blood samples from newborns, identifying newborns with SCD and introducing them to comprehensive care. The feasibility of the program has been mainly due to the approach of selecting public hospitals, integrating NBS services in existing government Reproductive and Child Health (RCH) services and potentially a shared population register.

We recommend that the NBS program is approved as a national program and should be coupled with an immunization program to leverage resources and increase feasibility. We recommend a stepwise scaling up from MNH and Temeke Regional Hospital to all peripheral hospitals in Dar es Salaam and subsequently to zonal hospitals across the country.

Authors’ contributions

JM, JK, HM, SEC, NU, AS and SR designed the study. LM, DS, JM, SN, VM, PM, PM, IM, MN, NU, SEC, GSK, MA, BSD, FT, ML, YD and FM implemented the study. BPM and SN analyzed and interpreted the data. SN and JM drafted the manuscript and all authors contributed to the drafts of the manuscript and approved the final draft.

Acknowledgements: Patients and staff at Muhimbili National Hospital and Temeke Regional Hospital, Ministry of Health, Community Development, Gender, Elderly and Children and Muhimbili University of Health and Allied Sciences: Muhsin Aboud, Henry Mwanyika, Said Aboud and Joyce Masalu, Dar es Salaam Medical officer (Grace Magembe), Joseph Eshun, COSTECH (Hassan Mshinda and Flora Ismail Tibazarwa).

Funding: This work was supported by the UKAid/Department for International Development (DFID) through the Human Development Innovation Fund (HDIF) [HW2.0003].

Competing interests: None declared.

Ethical approval: Ethical approval was granted by the MUHAS research and publications committee with reference number 2017-03-22/AEC/Vol.XII/67. Verbal consent was given by the participants (mothers of the newborns) of the study, which was recorded on the enrollment form. Verbal consent was formally approved by our ethics committee.

Author notes

Siana Nkya and Lillian Mtei contributed equally.