Availability of Safe Second-Trimester Abortion Services in Health Facilities in Accra, Ghana

We examined providers, methods employed, cost, and other determinants of availability of second-trimester abortion services in health facilities in Accra, Ghana in 2019 to inform policy and program decisions. A two-stage mixed quantitative and qualitative study designs were employed in the conduct of the study. The first stage was a short interaction of the mystery client with a clinical care provider to identify health facilities that provide second trimester induced abortion, the cost, and referral practices, where the facility did not have the service. The second stage was in-depth interviews of second-trimester abortion care providers and non-providers in various health facilities. For internal validity, it also explored the procedure cost, referral, and other practices at the health facilities included in the study, independent of what was captured in the mystery client survey. Second-trimester abortion services in Accra, Ghana are widely unavailable even in most facilities that provided abortion services. Referral policies and practices indicated by the service providers at various facility levels were inadequate. Criminalization of the procedure, social stigma, and fear of complications are the main factors that adversely influence the availability of second-trimester abortion in health facilities in Accra. Albeit increasing demand for second-trimester abortion in health facilities in Accra, services are not readily available due to the ambiguity of the law, its interpretation, and limited flow of accurate information on providers. Policies and programs that limit access to Second-trimester abortions in Ghana are amendable to ensure safe services. We examined providers, methods employed, cost, and other determinants of availability of second-trimester abortion services in health facilities in Accra, Ghana in 2019 to inform policy and program decisions. A two-stage mixed quantitative and qualitative study designs were employed in the conduct of the study. The first stage was a short interaction of the mystery client with a clinical care provider to identify health facilities that provide second trimester induced abortion, the cost, and referral practices, where the facility did not have the service. The second stage was in-depth interviews of second-trimester abortion care providers and non-providers in various health facilities. For internal validity, it also explored the procedure cost, referral, and other practices at the health facilities included in the study, independent of what was captured in the mystery client survey. Second-trimester abortion services in Accra, Ghana are widely unavailable even in most facilities that provided abortion services. Referral policies and practices indicated by the service providers at various facility levels were inadequate. Criminalization of the procedure, social stigma, and fear of complications are the main factors that adversely influence the availability of second-trimester abortion in health facilities in Accra. Albeit increasing demand for second-trimester abortion in health facilities in Accra, services are not readily available due to the ambiguity of the law, its interpretation, and limited flow of accurate information on providers. Policies and programs that limit access to Second-trimester abortions in Ghana are amendable to ensure safe services.


Background
Access to safe abortion care according to the World Health Organization in 1995 is a fundamental human right. As essential health care in most nations including Ghana, safe abortion services are expected to be available at high standards that are acceptable and accessible to all people within local jurisdictions. Traditionally, abortion care is divided into two: first trimester (week 1 through week 13) and second trimester (week 14 through week 27) of pregnancy (Rhoden, 1985). Induced abortion is a willful termination of pregnancy before the fetus becomes viable (before 28 weeks' gestation, which is approximately a fetus with birth weight less than 1000 g (WHO, 1995). Worldwide, about 210 million pregnancies occur annually of which an estimated 46 million end in abortions; (36 million in developing countries and 10 million in developed countries) (Shah & Åhman, 2004). Empirical evidence has shown that demand for second-trimester abortions is increasing across the globe. (Harries et al., 2012;Brookman-Amissah & Moyo, 2004;Bastianelli et al., 2014).
Evidence exists that second-trimester abortions are more often associated with procedural complications that could have serious consequences for the woman than first-trimester abortions. Yet, due to varied reasons, access to secondtrimester safe abortion services continues to be a challenge in many jurisdictions (Harris & Grossman, 2011). The reasons include varied legislations, (Jones & Weitz, 2009), cost of services, (Hodorogea & Comendant, 2010), inadequate flow of accurate information to service seekers and providers, (Lince-Deroche et al., 2015), social stigma (Puri et al., 2012), religious and moral values, and provider's willingness to provide second-trimester abortions (Gilligan et al., 1971;Dalvie, 2008;Kishen & Stedman, 2010).
In Ghana, empirical evidence on the incidence of second-trimester abortions is lacking. Anecdotal evidence has shown that demand for second-trimester induced abortions from health facilities in Ghana is increasing. For reasons ranging from stigmatizing providers and abortion seekers, lack of service delivery facilities to provider bias and negative attitudes (Aboagye et al., 2007;Turner et al., 2008;Martin et al., 2011;Appiah-Agyekum, 2014;Aniteye & Mayhew, 2013), There are limited published data on the availability of second-trimester abortion services to guide decision making to effectively address the increasing demand.
Induced abortion has been criminalized in Ghana since the development of the criminal code, Act 29 of 1960. However, Ghana relaxed her abortion law in 1985, to make abortion legal under broadly defined circumstances that culminated in the development of national policy, standards, and protocols to improve access to care to reduce abortionrelated maternal morbidity and mortality in the country. [20] Indeed, Ghana is one of the few countries in Sub-Saharan Africa with liberal abortion laws that permit induced abortion up to 28 weeks of gestation (Republic of Ghana, 1969;Singh et al., 2009). The national comprehensive abortion care policy, standards, and protocols try to optimize reproductive rights including allowing women to seek abortion care without a partner or spousal consent. (Ghana Health Service, 2012). However, in many situations, the decisionmaking process for the second trimester induced abortion is complicated by a number of factors that make services not readily available and accessible to many vulnerable women in need (Alex & Hammarström, 2004;Kjelsvik et al., 2018). Some of these factors are amenable to change to improve access to care or influence policy formulation and implementation to ensure availability and accessibility of highquality services to enhance the promotion of reproductive rights that are fundamental to national development.
In our study, we accessed the availability of secondtrimester induced abortion services in Accra Metropolis in 2019. Specifically, we sought to identify health facilities, care providers, methods employed, the cost, and other determinants of the availability of second trimester abortion care in Accra. Findings from the study would build local evidence that might inform the review of national and local policy and program decisions on the improvement of safe abortion care in Ghana. The findings would also be useful in planning further studies in this important but rarely investigated aspect of reproductive health in Ghana.

Contextual Framework
All over the world seeking and providing second-trimester induced abortions are more challenging than first-trimester abortions because of targeted legislation that imposes limits on gestational age that restrict abortion access in many jurisdictions (United Nations/ICPD, 1994). There are also moral dimensions to late or second trimester terminations with the use of terms such as "partial-birth abortion" and "born-alive abortion".
Since the landmark commitments made by various political leaders at the International Conference on Population and Development (ICPD) in Cairo, 1994, (United Nations/ ICPD, 1994UNFPA, 1994) issues relating to accessing safe induced abortion services remain public health and a reproductive right challenge among many nations worldwide (Berer, 2000;Morhee & Morhee, 2006;Rehnström Loi et al., 2015).Whilst many nations have successfully made policy and program decisions to minimize firsttrimester abortion-related deaths and illness, availability and access to second-trimester abortion remain inadequate amidst increasing demand for the service, particularly in the developing world (Singh et al., 2018).
A number of factors contribute to the demand for second-trimester abortion. These include law and policy environment, moral and cultural issues, health services infrastructure, human resources, and service delivery standards and protocols (Berer, 2000). Second-trimester abortions are riskier than first-trimester ones. For good outcomes, second-trimester abortions require a more comprehensive service delivery set-up, including blood transfusion, operation theatre with facilities for emergency surgery, referral, and transport, well-resourced referral destination. These are often not available in rural and less developed settings. The inadequate availability of the services has been compounded by the lack of reliable data on the demand and availability of second-trimester abortion services at both the national and international levels for comprehensive planning (Rehnström Loi et al., 2015).
In 2006, the Government of Ghana, in partnership with other organizations, launched the Reducing Maternal Mortality and Morbidity (R3M) program in seven districts within the Greater Accra, Ashanti, and Eastern regions, to improve comprehensive abortion care services (Sundaram et al., 2015), as permitted by the Ghanaian abortion law (Republic of Ghana, 1960). Ten years after the implementation of the R3M project and other interventions including advanced training of reproductive health fellows in Ghana, the nation has recorded a decline in abortion-related deaths and illness in the project implementation regions (London, 2015). Examining the availability and accessibility of safe second trimester induced abortion services across the various health care delivery sectors in Accra, Ghana, will enable policy makers to put in intervention to improve access to care.

Study Design
Mixed quantitative and qualitative study designs were employed in the conduct of the study. It was designed as a two-stage survey to capture quantitative and qualitative data. The first stage was a mystery client survey designed to be short interaction of the mystery client with a clinical care provider to identify health facilities that provide secondtrimester abortion and the cost of the services. Where the facility did not have the service, the survey was to assess the referral practice.
The second stage of the study was a qualitative survey of health care professionals (abortion care providers) in health facilities that had and those that did not have second-trimester abortion services. The qualitative phase of the study involved data collection from participants using mystery clients. Mystery clients' surveys have been reported as useful for such studies (Boyce et al., 2006;Sigdel et al., 2022). We therefore used this method to ensure that reliable data on the cost of service and referral practices are well documented as real as possible. The qualitative approach also explored reasons for provider choices regarding second-trimester abortion care practices at various health facilities studied.

Setting and Population
The study populations were healthcare professionals who provide clinical care for abortion care seekers at various health facilities that were well known for providing induced abortion services in Accra, Ghana.

Inclusion/ Exclusion Criteria
Facilities included in the study were public and private hospitals, private clinics, maternity homes, and reproductive health centers run by non-governmental organizations (NGOs) that are known in the local communities of Accra as abortion care providers. Other facilities included in the study were pharmacy and over-the-counter drug seller shops in the community with abortion care providers who sell and dispense medications used for induced abortion. Teaching hospitals were excluded from the study because they are, by policy, to provide only referral services which, do not include uncomplicated induced abortion care. Participation in the study was voluntary; health workers who were hostile and unwelcoming to the request for participation in the survey on abortion were left out of the study. referral practices, as well as reasons for second trimester abortion choices.

Data Collection Procedure
Data were collected between January and November 2019. With prior appointment, each selected health facility was visited by trained research assistants. At each selected facility a lead abortion care provider at post on the day of the visit was contacted and the purpose of the study and informed consent obtained after interviews are conducted. All the five field assistants were oriented on various provider behaviours and reactions to requesting for second-trimester induced abortion in a sensitive cultural environment before the fieldwork. All interviews were done face to face in English and each one lasted for 30-45 min.
In the first phase of the survey, the recruited mystery clients visited the selected clinics in pairs. They went through the facility reception and made a request to see the abortion provider. During the interaction, they made a request for the termination of pregnancy at about four to five months and wait for provider response regarding availability and cost.
If provider indicate non-availability, the mystery requests referral help. Responses are well noted and the data capture form completed soon after leaving the facility. On return from the field, the mystery clients had debriefing or interview sessions with the principal investigator when the mystery clients filed the data capture form with indication of observations and experiences in the facility visited. In the second stage, the interviews were conducted in private serene environments after assuring participants of confidentiality of the information collected and securing informed consent. The in-depth interviews were digitally recorded and field notes were also taken.

Data Analysis
The quantitative data captured were entered in Microsoft Excel and analyzed using SPSS version 18 using a semiquantitative approach and the results shown in appropriate tables. The recorded qualitative data were transcribed verbatim. The transcripts were thoroughly read several times to identify the main and sub-themes that were coded and recoded for thorough analysis. The findings were presented under sub-headings and supported with direct quotations from responses of participants, where applicable.

Ethical Considerations
The mystery clients used in the study participated voluntarily. The facilities visited and participants in the study were coded during data collection and processing to ensure

Sampling and Selection of Participants
No study has been found that estimated availability of second trimester induced abortion care in health facilities in Ghana. Hence, the proportion of abortion care provider health facilities that offer second-trimester induced abortion in a community in Ghana, the main objective of the study is not known. In sample size determination a snowballing approach was used to identify 155 facilities that provided induced abortion care within the community under study. Assuming 50% of the facilities provided second-trimester abortion care; at 80% power and 95% confidence interval of error, a sample size of 52 with 10% margin was estimated using StatCalc program of EpiInfo, version 7. From the list of 115 facilities identified to provide induced abortion services, 52 were randomly selected using the assigned serial numbers that were picked one after the other without replacement to constitute a set of facilities for the conduct of the survey.
For data collection, five females aged twenty to thirtyfive years and of varying backgrounds were purposively selected as mystery clients seeking second-trimester abortion care. They were given a one-day orientation including faking a pregnancy. Based on the outcome of the mystery client survey, a follow-up purposive sampling was done to select eighteen second-trimester abortion providers from the identified facilities providing second-trimester abortions to solicit their views on demand and availability for secondtrimester abortion services in their respective facilities. For more balanced views, abortion care providers care from health facilities where second-trimester abortions are not available were also selected to participate in the qualitative survey. Recruitment continued to point of saturation. Health care professionals included in the survey were medical officers and midwives working in public and private health facilities, pharmacists, and chemists. Only health care facilities with abortion care providers who voluntarily and willingly offered to be interviewed to solicit their views on second-trimester abortion services in their respective facilities were recruited to participate in the in-depth interview.

Data Collection Tool
Both qualitative and semi-quantitative data were collected using an interview guide and a questionnaire respectively. The data collection instruments were developed from literature review by the authors. The questionnaire captured data on facility type, availability of second-trimester abortion, cost of second trimester abortion, and type of referral offered. The interview guide focused on facility type, kind of provider, level of training, years of experience/service, second trimester abortion practices including methods used, of 10 pharmacy and chemical seller shops surveyed provided second trimester abortion care (Table 1). In summary, whereas second-trimester abortion care was not available in public health hospitals, some private hospitals, clinics, pharmacy and chemical shops provided the service. Table 2 shows the background characteristics of respondents who participated in the in-depth interviews. With respect to healthcare professionals who provided second-trimester terminations, an in-depth interview of 18 providers and 5 non-providers of second-trimester abortion revealed that physicians, midwives, pharmacists, and over the counter drug sellers were involved.

Types of Healthcare Professionals Providing Second-Trimester Abortion
In the second stage of the study, three medical officers, twelve midwives, two pharmacists, and two over-the-counter drug seller shops from identified facilities that indicated anonymity. The Ghana Health Service gave ethical approval for the study (GHS-ERC: 02/09/2016). Participation was free and voluntary, and respondents were allowed to decline the interview at any time. Considering the sensitive nature of abortion and the study objectives, only providers or key informants who were receptive to the request for participation were engaged in the in-depth interviews. Table 1 shows findings from the mystery survey of the 50 health facilities included in the study. The facilities studied comprised five (5) reproductive centers (clinics) run by NGOs, seven (7) public hospitals, three (3) quasi-governmental hospitals, nine (9) private hospitals, eleven (11) maternity homes, five (5) private clinics, five (5) pharmacy and five (5) over-the-counter drug seller shops. Thus, 19 hospitals and 21 clinics and 10 pharmacy or chemical seller shops were studied in the first stage.

Type of Healthcare Facility Providing Second Trimester Abortion
Of the 50 health facilities explored, our mystery abortion care seekers identified 18 facilities that provided second-trimester abortions whilst 32 did not. All abortion care providers in the ten public hospitals including quasi-governmental hospitals indicated not providing second-trimester abortion care while four out of nine private hospitals provided the service. Regarding clinics, all providers from reproductive health centers (clinics) run by non-governmental organizations stated not providing second-trimester abortions while two out of five and six of 11 from private clinics and maternity homes provided the service respectively. Similarly, six

Legal and Policy Concerns
First and second-trimester terminations are not differentiated in the Criminal Offences Act and mention is only made of termination before the period of gestation is completed. The results show that the lack of knowledge of the legally acceptable gestational limit and fear of legal consequences in providing a second-trimester abortion are factors influencing availability of second-trimester abortion services. Some midwives were of the opinion that they have been trained to perform abortion up to 12 weeks only ……we always see them but unfortunately we can't help since the facility only, has midwives who have been trained to provide abortion services only up to 12 weeks here (Midwife).
The providers were not sure of the legality of providing second-trimester abortions in Ghana. Some think that by providing second-trimester abortion the provider could face disciplinary action that could result in withdrawal of their professional license to practice or even face prosecution with potential imprisonment. Consistent with this opinion, a midwife working at a reproductive health center run by a non-governmental organization also indicated…I will not risk my professional licenses or go to jail by trying second-trimester abortion although I know how to do it and can even do it better than a gynecologist (midwife, NGO facility).
Some providers, though aware of the legal and professional limitation and potential legal consequence, they however exhibited some determination to provide the service: I know what I'm doing is illegal though, but I only sell the pill out to people that I believe will not put me into any trouble (over-the-counter drug seller). Another provider indicated: Generally, we in the government hospitals provide only first-trimester abortion at the family planning unit…but after options counseling fails, I risk to induce them and ask them to return when bleeding starts……. (Midwife, public hospital).
Based on providers' level of understanding and interpretation of the abortion law, some participants used the law to explain their non-performance of second-trimester abortion practices; …this facility provides abortion services as mandated by the law. We believe that, although abortion is legal in Ghana, the law frowns on providing abortion above 12 weeks of gestation hence, we do not perform such services (Midwife, NGO facility). Similarly, another midwife indicated: the abortion law does not permit my facility to go above 12 weeks so I will never do anything that will send me to jail… (Midwife, maternity home).
Some providers although aware of professional and other limitations, think There are a need to assist abortion care seekers; …we are not supposed to stock or sell abortion drugs in the chemical shop, but sometimes the people we to be providing second trimester abortions. The 23 various cadres of abortion care providers who voluntarily and willingly offered to be interviewed to solicit their views on demand and clients' reasons for second-trimester abortion practices in their respective facilities were included in the in-depth interview.

Demand for Second Trimester Abortion Care
The study participants indicated the existence of high and desperate demand for second-trimester terminations by a range of women in need. It's a pity many women in Ghana are looking for a place to have an abortion at advanced gestations (midwife, maternity home); …different kinds of people come to us looking for some drugs for an abortion (Pharmacists). Similarly, a private physician and a midwife indicated; daily people walk in here requesting second-trimester termination (Medical officer, Private hospital); since the demand keeps increasing, we are considering starting soon (Midwife, Private hospital).
The findings further indicated that second-trimester abortion care seekers show features of stress and desperation at various healthcare facilities where they presented to find a solution to the unwanted pregnancy. …They always come crying that they didn't know they were pregnant until this late (midwife, public hospital).
The choice of provider could be influenced by the premium on offering the service. Some may not recognize the exigency to help provide the needed care. Yes, women come here requesting second-trimester abortion services, but, providing second-trimester abortion services is not our priority for now since we are very busy with other services (Medical officer, Quasi-governmental Hospital).
Desperation and helplessness are associated with a need for abortion. Some participants indicated concerns about the worry and extreme anxiety that the abortion care seekers exhibit when they visit their facilities; Some are so desperate that if we tell them we can't help them, their mood changes and some even start to cry like babies begging us to help them at all cost, so we sell the abortion pills to them. (Pharmacists)

Factors Influencing the Availability of Second-Trimester Abortion Services
A number of factors influencing the availability of secondtrimester abortion services in Accra metropolis were identified as follows. and Ghs800.00 (65usd) while the Clinics charge procedure fees between GHS 2000.00 (616usd) and GHS 2500.00 (202usd). Pharmacy and chemical seller shops were the least expensive health facility to seek the second-trimester termination; these facilities charged between GHS 300.00 (24usd) and GHS 600.00 (48usd).
Safety, expected cost of the procedure, perceived socioeconomic status of the client, and her ability to pay, were important considerations of providers in making choices for referral of care seekers that present at their health facilities; I would have preferred referring my clients to a place where they can have a safe abortion at a cheaper cost, but the hospitals in this area are very expensive; because they can't afford their services, I sell the abortion pills to them at a cheaper price (Chemical shop attendant).

Methods of Abortion and Cadre of Providers in Health Facilities
The provider cadre and methods used for the termination was a key determinant in accessing a second-trimester abortion from a facility. Although we observed a mixed method of abortion with a varying cadre of providers across the various health facilities ( Table 2), There were a general misconception that medication abortions were only available in the pharmacist and Over-the-counter drug sellers' shops.
Respondents were of the view that to avoid the surgical method of termination which is perceived invasive and for that matter more dangerous, some clients would opt to visit the pharmacist or Over-the-counter drug sellers for medication abortion. A respondent explained that……the majority of our clients express fear of D&C hence they come to us requesting for medications to have a termination…. (Overthe-counter drug seller).

Health Facility Infrastructural Need
The need for good infrastructure support for inpatient care has been clearly indicated by the trained providers in the study …unlike abortion in the first trimester, we usually admit our clients to the hospital during second-trimester termination and discharge them home only when we are very sure they are fit to go (Midwife, maternity home).
Some participants stated that, the nature and set up of the health facility sometimes limit the capacity and the ability of seemingly competent providers to handle second-trimester abortions: we have two locum medical officers who provide specialist obstetrics and gynecological services including abortion services weekly when we book clients. They have advised us to expand our facility and equip it to run 24-hour services so they can introduce second-trimester abortion services (Midwife, Private hospital). see here are so helpless that I have just kept a few stocks to help people who are very desperate so that, they don't end up going to the herbalist who will destroy their womb or kill them with herbal concoctions…. (Over-the-counter drug seller).

Moral Values
Moral values and the stigma associated with second-trimester abortion featured prominently during the provider interview. Some providers perform the procedure secretly; … I know some doctors secretly do bigger gestations in the theater and at their private facilities to avoid stigma… (midwife, public hospital). Stigma may arise from unexpected complications, -----my boss encountered serious complications in the past that dented his image in the community, hence has decided not to invest in second-trimester abortions any longer (Medical officer, Private hospital). Some forced their values on clients: ------I don't believe in providing second-trimester abortion; so I advise them to give birth and sometimes also put fear in them so they don't do it (Midwife, public facility).

Safety of Second-Trimester Terminations
There was a general fear of complications associated with second-trimester terminations throughout the interview. For instance, -------we have a big and well-equipped facility here that provides specialist obstetrics and gynecological services, but we limit our abortion services to the first-trimester because it's safer. (

Cost as a Deciding Factor
The study has established cost as a deciding factor for accessing second-trimester abortion in 18 of 20 health facilities as shown in Table 1. The table also shows the second-trimester abortion methods available at various health facilities and the corresponding cost. While hospitals, clinics and maternity homes provide both medication and surgical second-trimester abortion care, pharmacy and over-the-counter drug seller shops, herein referred to as chemical shops provide medication only method. The cost varied with the type of health facility and the healthcare professional providing the service.
For hospitals in the study, the cost of a second-trimester abortion in Ghana cedis (GHS) was between GHS 1000.00 (81usd) and GHS 3000.00 (242usd). Most maternity homes charged procedure fee of GHS 500.00 (40usd)

Poor Treatment vs. Referral and Care-Seeking Facility Choices
Some study participants, particularly those from private clinics were of the opinion that abortion care seekers do not receive fair treatment in public hospitals. Thus, midlevel providers prefer private to public hospitals in referring their clients who need hospital or physician attention. In explaining referral facility choice, a midwife indicated that, '……we refer them to a private hospital that we collaborate with to help because the public hospitals that we know don't treat our clients well as most of the time they humiliate them and drive them away to go and give birth' (Midwife, NGO facility).

Awareness of Limitations and Referral
Abortion care providers are aware of the potential complication of second-trimester abortion but have different approaches to mitigating the challenge. Some are not willing to assist second-trimester abortion care seeker even with referrals. ……I limit myself to only the first trimester; I don't refer clients seeking second-trimester abortion services to avoid possible complications they may suffer (Midwife, public facility).
Some held contrary views and are more sympathetic. …although I have a conscientious objection to second-trimester abortion, I believe that if we don't do it, some quacks will do it unsafely and the complications will come back to me so 1 try to refer the clients to a place where I'm sure they can have safe services'(Midwife, maternity home). Another provider indicated '…we refer our clients to sister private hospitals where we know the service is provided' (Midwife, Private hospital).
Some think the services should be limited to physicians ---'. Second-trimester abortion should only be done in a specialist facility and by a trained medical officer; so, it is very dangerous and criminal for it to be done outside a wellresourced hospital by a non-medical practitioner' (Medical officer, private hospital).

Discussion
Safe second-trimester abortion services are scarcely available within health facilities in Accra metropolis, as most of the health facilities that offer second-trimester abortion service are privately owned clinics, maternity homes, pharmacies, and chemical seller shops that are not mandated by the abortion policies and laws.
The cost of services varied with age of gestation, and level of facility, and cadre of provider. It ranged between

Human Resource Challenges
The provider skills and lack of support was another concern as a respondent indicated: -----'my former medical director who used to support us anytime There are a complication or police case during service delivery is no more and those in-charge now don't even care much about us, so you are on your own if something happens' (midwife, NGO facility).

Type of Provider
Whereas the medical officers, nurses, and midwives were reported providing second-trimester abortions in a clinical environment, the pharmacists and chemical shop attendants dispense abortion pills 'over-the-counter' to very desperate clients and those whom they perceive as not spies on them. A respondent indicated that: ---Just to help our clients, I sell some medications to them and give them directions on how it should be used when they get home……. It is a very risky thing to do because some of them bleed badly and they end up in the hospital…… (Pharmacists).
Providers indicate the need for teamwork in carrying out second-trimester termination: Our second-trimester abortion services are initiated by our specialist and the nurses are asked to monitor the client until the pregnancy is termi-nated…. Sometimes we the nurses do everything and only call in the doctors if There are a complication. (Nurse, private hospital).

Missed Opportunity
There were some missed opportunities to integrate secondtrimester abortion services into the service mix at some health facilities. ------'We have a big and well-equipped facility here that provides specialist obstetrics and gynecological services including first-trimester abortion, but my boss, does not want to hear anything relating to secondtrimester terminations…….and will not invest in it' (Medical officer, public hospital).

Referral Systems for Second-Trimester Abortion
Various referral systems were observed across the facilities that do not offer second-trimester abortion services. Predominantly the private facilities were the main destinations for the referrals (Table 4).
There were many reasons/justifications for the private facilities being the key destinations for the second-trimester abortion. Some of which includes: were noted in the identified facilities that provide secondtrimester terminations. The observation that regardless, of the facility type, midwives constitute the majority of providers is in line with the task-sharing policy of the country but not consistent with the national protocols. The continued use of services of pharmacists and chemical shop attendants who are legally not permitted to provide abortion care suggests inadequate access to appropriately trained providers in the study area.
The Ghanaian abortion law (Republic of Ghana, 1960), and various abortion care standards and protocols (Ministry of Health, 2003, Ghana Health Service, 2012, 2014, clearly state the cadre of health workers mandated to provide safe abortion services in Ghana. These legal and policy provisions also stipulate the environment considered safe for an abortion service to be done. Basing this framework on the World Health Organization's (WHO) definition of unsafe abortion (i.e. 'a procedure for terminating a pregnancy performed by persons lacking the necessary skills or in an environment, not in conformity with minimal medical standards, or both) (WHO, 2012) and 1992 WHO Technical Consultation report, (WHO, 2008).
The observation made so far could be interpreted as second-trimester abortion seekers are receiving unsafe abortions from pharmacies, chemical shops, and private clinics operated by pharmacists, chemical shop attendants and nurses respectively although WHO guidelines have recommended medication abortion using mifepristone and misoprostol or misoprostol alone as safe (WHO, 2015). The observation that abortion services although done undercover are available in chemical and pharmacy shops at no gestational limits, makes the outcome of second-trimester induced abortions difficult to measure since they might be underreported or misclassified by both the abortion seeker and provider at the community and facility levels.
The findings that the pharmacists and the chemical seller shops did not indicate a limit of gestational age reveals the deficiency in the policy, standards, and protocols that excluded them from training in abortion care provision. Clearly, the midwives recognize their professional limitation whereas the untrained chemical seller provides the procedure at any gestation although not completely oblivious of the dangers and the consequences of their actions. We hereby indicate that the inclusion and training of other health care providers in the national CAC protocol will not only increase access to care but minimize the incidence and severity of complications. For such situations, key peripheral care would be able to make timely referrals to public and private hospitals and clinics without fear of prosecution that is rarely done anyway. Additionally, their training will promote the flow of accurate information which is fundamental to improving access to self-managed abortion the GHS 300 in over-the-counter facilities and GHS 3,000 in private hospitals and clinics. Obviously, the level of the cost of second trimester abortions could be beyond the affordability of the average woman in Ghana, particularly poor rural women who may be in need under varied circumstances and found in the study. The findings from our study are consistent with the narrations of chemical shop attendants in a previous study, where high costs of abortion were the main reason for the helplessness of some women seeking care at their facilities. The implications of our findings is that previous interventions by the government of Ghana and development partners to improve access to safe abortion services have not had the desired effect, [19,31,33,34] hence policy makers and reproductive health programs implementers need to consider other initiatives for increasing access to safe abortion care including second-trimester services as permitted by Ghanaian law.
The various factors identified to be associated with the limited availability of second-trimester abortion services in Ghana remain the same as those that hitherto restricted access to safe abortion services before previous interventions including the introduction of the R3M and international family planning fellowship projects. As found in our study, accessing safe second-trimester abortions has been linked to inadequate knowledge of the law including gestational age limit and fear of legal consequences of providing a second-trimester abortion in Ghana. Other factors found in the study; social stigma, local policy of health facilities, provider moral values on the provision of second-trimester abortion are well documented factors that influence access to abortion care locally and internationally. Similarly, the nature of the facility, the clinical setup, provider training and skills, and general fear of complications associated with second-trimester terminations are recognized factors that influence access and outcome of abortion as clinical care. Nonetheless, most of these factors are changeable to improve access to safe second-trimester abortion.
The persistence of these factors has over the years prevented even legally qualified competent clinicians to provide safe second-trimester abortion services in Ghana. Consequently, creating the enabling environment for safe induced abortion to avert abortion-related deaths and injuries remains a public health crisis that violates reproductive rights including the dignity of women. Many women and girls in Ghana are compelled to carry pregnancies to term against their will because they lack access to second-trimester termination or could not afford the high cost of such services even where they are available or resort to unsafe abortion providers (Chemical sellers and pharmacist) for support.
Various cadre of service providers (Medical officers, Midwives, Pharmacists, Nurses, and Chemical shop attendants) inputs. All three authors read and approved the manuscript before submission for publication.
Funding Not applicable.

Data Availability
The raw data and any material related to the study are available upon reasonable request from the corresponding author.

Declarations
Ethics approval and consent to participate All participants gave both written and verbal consent to participate in the study. The Ghana Health Service gave ethical approval for the study (GHS-ERC: 02/09/2016).

Consent for publication
All participants gave both written and verbal consents individually for the publication of the research findings on one condition that their personal and institutional identities remain anonymous.
most likely global approach to reducing the impact of social stigma and other accessibility factors that negatively affect outcomes of unwanted pregnancy management. However, legal and moral challenges need to be explored in comprehensive local stakeholder surveys, considering the limited availability of research work on second-trimester abortions.

Study Limitations
The study encountered two main limitations. The first was in relation to selection of participants. Being a sensitive study, majority of potential participants refused to participate in the study because they did not want to be associated with abortion. This challenge delayed data collection. The second limitation was lack of funding. Being a self-funded project with limited funds, the high cost of consultation in some facilities also prevented us from going beyond the initial interaction at the OPD to get more information on what actually happens to clients requesting for second trimester abortions beyond the OPD stage. In all these we managed to collect as many relevant data as possible based on which the results are presented. In the light of these limitations, particularly the sample size, our findings cannot be generalised but could provide some base line information to guide a bigger study in future.

Conclusion
Despite the high demand, safe second-trimester abortion services including referral are not widely available in wellknown hospitals that provide abortion care in Accra. The lack of a reliable referral system for second-trimester abortion poses a public health challenge to reducing abortion related maternal morbidity and mortality. Findings from the study calls for a strategic planning towards public health education on the abortion law and the availability of safe abortion services to reduce the need for late abortion. The Ghana Health Services should also equip its facilities and build adequate human resource capacities for ensuring the provision of safe second-trimester abortion services when necessary. More research is required to explore why women delay till the second-trimester before seeking safe abortion care.