Cost analysis and service establishment of in-house amniotic membrane transplantation at a tertiary hospital in South Africa

¹University of Cape Town, Groote Schuur Hospital Ophthalmology, Cape Town, South Africa

^&Corresponding author
Jonathan Timothy Oettlé, University of Cape Town, Groote Schuur Hospital Ophthalmology, Cape Town, South Africa

Introduction    

Many conditions lead to corneal scarring, with varying prevalence across populations. In low and middle-income countries, infective keratitis is a leading cause of monocular blindness [1], while chemical injuries account for 11.5-22.1% of global ocular trauma cases [2]. Management typically includes a combination of medical and surgical interventions. Recently, the use of the human amniotic membrane (AM) on the ocular surface has gained popularity. It can be applied topically using a scleral contact lens, mounted on a ring conformer, or attached with sutures or tissue glue. Amniotic membrane (AM) has been successfully used in numerous ophthalmic conditions and is generally regarded as safe and well-tolerated [3]. South Africa experiences a significant burden of corneal pathology, particularly within the state healthcare sector, which lacks adequate access to AM. This study provides a cost analysis of in-house AM tissue recovery and outlines the process of establishing such a service. Amniotic membrane (AM) comprises an epithelial cell layer, basement membrane, and connective tissue stroma. Its use in medicine dates back to 1910 [4], with the first ophthalmic application by de Rotth in 1940 for conjunctival reconstruction [5]. Currently, AM is used in treating conditions such as corneal burns, Stevens-Johnson syndrome, persistent epithelial defects, and neurotrophic keratopathy [6,7].

The basement membrane of AM resembles that of the conjunctiva and cornea [8], providing a scaffold for re-epithelialization. It supports a moist ocular surface, protects against mechanical irritation from the eyelid, and promotes epithelial cell adhesion and proliferation [8-10]. AM also inhibits fibroblast proliferation and myofibroblast transformation, thereby encouraging wound healing and reducing scar formation by suppressing Transforming Growth Factor B [11]. Its anti-inflammatory effects result from the inhibition of pro-inflammatory cytokines and Matrix Metalloproteinases [12,13], and its overall effect appears to be anti-angiogenic [14]. In addition, AM contains several antimicrobial molecules [6] and does not express Major Histocompatibility Complex class 2 antigens [15], making it suitable for transplantation without the need for immunosuppression [16]. There is no risk of tumorigenicity[8,17], and no communicable disease transmission has been reported from ophthalmic AM transplantation. Amniotic membrane (AM) is recovered post-delivery [18]. Preservation methods include cryopreservation, lyophilization, and air drying, although unpreserved frozen AM has better healing properties [14]. The rationale behind this study is to address the high cost and limited availability of AM in South Africa's state healthcare sector. Given the prohibitive cost of commercial AM in South Africa, in-house recovery offers a cost-effective alternative and could make AM accessible to more state healthcare patients. This study aimed to compare costs of in-house AM recovery with commercially available tissue, and to describe the process followed in establishing the service.

Methods     

Study design: in this observational study, we conducted a CMA using bottom-up microcosting for a newly established AM transplant service at a tertiary state hospital in South Africa. We analysed the index recovery event and the 1^st subsequent recovery event. Bottom-up microcosting involves determining the cost of each component of the service or intervention under study, including waste disposal and storage expenses. The process followed when establishing the service was recorded and described for other ophthalmic centres in resource-limited settings to be able to use as a blueprint. We also established a standard operating procedure (SOP) in consultation with Infection prevention and control services at our hospital, which was ratified by hospital management.

Setting: tissue recovery was performed at Groote Schuur Hospital, a tertiary-level hospital in Cape Town, South Africa, in 2023 and 2024.

Participants: the donors were expectant mothers identified by the obstetric service with no contraindications to donation and were scheduled for elective caesarean section for maternal indications. The tissue recovery was performed by a single recovery technician who was trained in the recovery of the amniotic membrane.

Variables: the size of the amniotic membrane varies slightly between individuals, leading to slight differences in the number of pieces recovered, and thus the relative price per piece.

Data sources/measurement: data was obtained directly from the relevant department: data on drug costs was obtained from the pharmacy, laboratory testing from the National Health Laboratory Service, and consumables from theatre nursing management, hospital stores and waste disposal. Historical data was obtained from purchase records.

Donor selection: eligible donors are expecting mothers scheduled for elective caesarean section. Donors are screened using a questionnaire and are categorised as “low risk” or “high risk” based on screening criteria - if any of the criteria is positive, the patient is “high risk” and is excluded from donation (Table 1).

Ethical considerations: donor informed consent was facilitated by the recovery technician prior to delivery, in a language of the patient's choice. The decision to donate was voluntary, did not affect the patients' care in any way, and no reimbursement was offered for the donation. Ethical approval for this study was granted by the Human Research Ethics Committee of the University of Cape Town, HREC Ref 519/2023.

Results     

Cost analysis results: as of March 2023, commercially purchased AM cost the state USD 252.58 for a 2x2 cm piece of cryopreserved tissue. Groote Schuur Hospital purchases between 10 and 15 pieces of AM annually, with the average annual expenditure ranging from USD 2,525.94 to USD 3,788.91. Each in-house tissue recovery costs the state USD 60.89 and yields between 75 and 100 pieces of 2x2 cm or larger tissue. The first tissue recovery yielded 81 pieces, while the second yielded 98 pieces. For this cost analysis, we based the calculations on the mean of these two events. With an average of two tissue recovery events per year, this results in direct cost savings of between USD 2,465.05 and USD 3,728.02 annually. Additionally, the hospital gains 140 to 180 extra pieces of tissue available for transplantation each year. The average cost per piece of tissue from in-house recovery is approximately USD 0.68, 0.3% of the cost of commercially available tissue. A detailed cost breakdown is provided in Table 2.

Establishment of the service: the first step in the tissue recovery process is registering the technician with a certified tissue bank for AM recovery. Once training and registration were completed, we presented a proposal to the hospital management board, following local policy guidelines for implementation. Figure 1 provides a flowchart detailing the process we followed to establish the service at Groote Schuur Hospital. Although specifics may vary between hospitals, the essential steps are outlined below. The process culminated in a Memorandum of Understanding (MoU) between the hospital and the tissue bank, providing a governing legal framework for the recovery process. The tissue recovery procedure itself is straightforward, performed under sterile conditions in the operating theatre. The Standard Operating Procedure (SOP) is linked in the references [19]. After receiving the placenta in theatre, cord blood is collected for infectious disease screening, and a pus swab is taken for culture. The AM is then separated from the underlying chorion by making an incision at the base of the cord with a No. 15 blade, followed by blunt dissection - video linked in the references [20]. The tissue is washed and decontaminated in multiple antibiotic and antifungal baths until thoroughly cleaned (Figure 2). During this process, adherent collagenous stroma is removed using a sterile swab between washes (Figure 3). The tissue's orientation is confirmed using gauze or a spear swab—epithelial tissue will not adhere to the swab, but the collagenous stroma will adhere in thin strands. Once clean, the tissue is mounted on sterilized tinfoil, cut to the desired size, and immediately frozen in sterile packaging (Figure 4). After infectious disease screening results are confirmed, the tissue is ready for use.

Discussion     

The primary objective of this study was to assess the cost-effectiveness of developing an in-house AM transplantation service within a tertiary hospital setting, particularly in resource-limited environments. Our secondary objective was to document the methodology followed in establishing the service and to provide a framework that can assist other medical institutions wishing to implement a similar service. Our analysis demonstrates a significant cost reduction when utilizing locally recovered AM compared to commercially available cryopreserved alternatives, with the latter being significantly more expensive. During the implementation of this tissue transplant service, we found no major barriers to establishing the service, and have documented the steps to be followed. The tissue processing itself does not require any specialized equipment and takes a single technician a few hours to perform.

Cost: amniotic membrane finds various applications in ophthalmology, but for many ophthalmic centres in low- and middle-income countries, cost poses a significant barrier. We have demonstrated the substantial cost advantage of establishing an in-house AM transplant service, with our service costing only a fraction of what we previously paid for commercially available amnion - commercially available cryopreserved AM is over 300 times more expensive than fresh-frozen tissue. A substantial portion of the total tissue recovery cost results from the nucleic acid tests for hepatitis and HIV; these tests are run three times each and are performed by our regional blood transfusion service, who perform the same tests on all human tissue, blood products, and solid organ transplant donors.

The following costs were not included: 1) Salaries for employees involved in the recovery process: as the recovery technician is a full-time hospital employee, adding tissue recovery to their regular duties and responsibilities did not result in additional costs for the employer; 2) purchase costs for reusable equipment: the impact of the tissue recovery on the lifespan of existing equipment is negligible, therefore purchase costs were not included. No new equipment was purchased for establishing the service; 3) autoclaving costs: our hospital theatre autoclaves run continuously during working hours, and adding items to a cycle does not incur extra costs. The items for sterilization (tinfoil, sterile self-seal pouches, scissors, forceps, and bowls) were incorporated into existing cycles without consuming additional resources; 4) costs related to service interruption: other ophthalmic services provided to our patients were not affected by the tissue recovery; 5) the costs of routine maternal 1^st-trimester screening tests: these form part of national guidelines for antenatal care.

The impact of having fewer staff members available can be mitigated if the department is aware of the recovery date in advance. The tissue recovery process takes one technician approximately 3 hours and occurs once every 4-6 months, making it feasible for most centres to manage the additional workload without disrupting regular services. It is noted that the surface area of the foetal membranes varies, with an average surface area of 1115 ±149 cm² [21]. We expect about a 15% variability in the size of the AM recovered, with an equivalent variability in the number of pieces retrieved and the relative cost per piece.

Service establishment: the process of establishing a novel transplant service in a tertiary state hospital will vary greatly between centres - in unregulated environments, establishing such a service would be considerably easier. Groote Schuur Hospital adheres to best practice standards in all aspects of its business policies; with dedicated and engaged teams in hospital management, ICU/transplant, O&G, ophthalmology, theatre management, and others, we were able to establish the service without any major barriers to implementation. The registration of the recovery technician with a tissue bank is critical for ensuring the legality of the process, which would otherwise be challenging to defend in the unlikely event of a vision-threatening complication. Two important documents used in establishing the service are the SOP and the Service Plan - the Service Plan details ´WHAT´ and ´WHY´, the SOP explains ´HOW´. Both of these documents need to be ratified by hospital management. Our informed consent document includes a patient info section and high-risk questionnaire and also requires approval from hospital management.

Infectious disease risk mitigation: proper donor selection and correct handling and preparation of the tissue have rendered the risk of communicable disease transmission with AM negligible. While a theoretical risk exists, to date there has never been a case report of communicable disease transmission from ophthalmic use of AM. Nevertheless, all reasonable precautions are taken to minimize this risk.

Steps taken to mitigate the risk of allogenic infectious disease transmission: 1) Initial maternal infectious disease screening in 1^st trimester of pregnancy, including HIV and syphilis. Only negative donors will be considered; 2) exclusion of high-risk donors - as outlined in Table 1; 3) repeat blood tests (cord blood) at the time of recovery: nucleic acid testing for HIV, Hepatitis B&C, and serology for syphilis; 4) a pus swab and/or tissue sample of the amnion taken at the time of recovery. If any pathogenic bacteria is detected, the sample is discarded; 5) the specimen is washed and decontaminated in multiple baths of antibiotic and antifungal solution, with doses and techniques elaborated below; 6) strict sterile procedure is followed, and performed in theatre; 7) immediate processing at the time of delivery - no prolonged storage or extensive handling before preparation; 8) immediate freezing of the tissue after preparation; 9) the remaining tissue is discarded after 6 months of storage.

Implications: the establishment of an in-house AM transplant service has significant implications for healthcare delivery in resource-limited settings, particularly in South Africa. The marked reduction in cost presents a compelling case for other state hospitals to adopt similar practices. This service not only helps allocate scarce resources more efficiently, but also broadens patient access to advanced ophthalmic interventions that would otherwise be prohibitively expensive. Furthermore, the simplicity of the recovery process, which does not require specialized equipment or substantial additional workforce investment, makes it highly replicable. If widely implemented, this model could make AM more widely available for ophthalmological interventions, particularly for conditions such as corneal burns and neurotrophic keratopathy, where its use is well-established. This approach could serve as a blueprint for state hospitals in South Africa and other resource-limited countries, offering a scalable, cost-effective solution to address both the scarcity of tissue and the financial barriers to access.

Limitations include placental size variation: smaller placentas could yield 10-15% fewer pieces, and would be less cost-effective. Additional recovery events would account for this variability; our cost analysis was limited to two recovery events. Additionally, we limited our analysis to direct healthcare-associated costs; other costs, including intangible costs, were not considered, and the healthcare cost saving effected by the additional availability of AM was not taken into consideration. The legal framework around which we established the tissue transplant service is South African; other countries will have different laws and regulations.

Conclusion     

In-house AM recovery at Groote Schuur Hospital has demonstrated a substantial cost reduction and significantly increased tissue availability. Our analysis revealed that in-house AM recovery costs approximately USD 0.68 per piece, compared to over USD 252 for commercially available tissue, achieving cost savings between USD 2,465 and USD 3,728 annually. Additionally, this approach yields 140 to 180 extra pieces of tissue annually. The process is straightforward, does not require specialized equipment, and is easily replicable in other settings. This cost-effective model improves access to AM, furthering ophthalmic care for patients in resource-limited environments. Future efforts should focus on broader implementation and continuous refinement of recovery practices to maximize the benefits of this approach for both local and regional healthcare systems.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Conception and study design: Jonathan Timothy Oettlé and Nagib Du Toit. Data collection: Jonathan Timothy Oettlé. Data analysis and interpretation: Nagib Du Toit and Jonathan Timothy Oettlé. Manuscript drafting: Jonathan Timothy Oettlé. Manuscript revision: Jonathan Timothy Oettlé and Nagib Du Toit. Guarantor of the study: Jonathan Timothy Oettlé and Nagib Du Toit. All the authors read and approved the final version of this manuscript.

Tables and figures     

Table 1: high risk screening criteria

Table 2: breakdown of cost analysis

Figure 1: flowchart detailing the administrative process followed in establishing the service

Figure 2: washing the tissue - note the collagenous stroma (arrow) separating from the membrane

Figure 3: cleaning with a gauze swab

Figure 4: packaging in sterile packets

References