Selecting an Automated Parenteral Nutrition Compounder Device at a Tertiary Care Institute in Saudi Arabia

Abstract

Background: With advances in hospital automation, selecting an automated compounding device (ACD) for compounding parenteral nutrition (PN) may be challenging. A well-designed comparative review that considers safety and technical standards could assist in the selection process. United States Pharmacopeia chapter <797> and The American Society of Hospital Pharmacists (ASHP) has provided guidance on the safety and accuracy of ACDs. Objective: Utilizing a checklist of technical and performance specifications investigators sought to assess the adequacy of compounders available in the Kingdom of Saudi Arabia. Method: Investigators compared the NutriMiX KS-240 and the Exactamix EM 2400. A literature review of practice standards, safety, technical, and performance specifications was conducted. These were developed into a checklist of 46 technical and performance specifications. Investigators evaluated each ACD using the checklist. Results: It was found that 43 (93.48%) of the technical and performance specifications, required by the hospital, were met by the NutriMix KS-240, while 39 (84.78%) were met by Exactamix EM 2400. Conclusion: This review found that although ACDs comply with the majority of standards and specifications, software integration with the hospital information system may be a limiting factor in selecting an ACD. The investigators concluded that the NutriMiX KS- 240 matches the technical and performance specifications for automated compounders required by our institution.

Introduction

Pharmacy automation has become a lucrative industry within the Kingdom of Saudi Arabia (KSA). In a national survey, Alomi et al found that 70% of participating hospitals, at the Saudi Arabian Ministry of Health uses automated parenteral nutrition compounding devices. ¹ The Ministry of National Guard Health Affairs provides free health care to members of the National Guard and their dependents. It is committed to safe patient care and has embraced the international standards of health care by adopting initiatives such as hospital automation. These include: computerized physician order entry, hospital integration, automated dispensing cabinets, carousels, and automated packagers. Parental nutrition (PN) at Prince Muhammed bin Abdul Aziz Hospital, Al-Madina Al-Munawarrah, was traditionally compounded by gravity filling of PN bags. To improve safety and efficiency, our institution sought to procure an automated compounding device (ACD) for PN. The United States Pharmacopeia chapter <797> standards provide guidelines safeguarding the accuracy and precision of ACDs. ² The American Society of Hospital Pharmacists (ASHP) guidelines on the safe use of ACDs for the preparation of parenteral nutrition admixtures provides well-established guidance in promoting standards, consistency, and safety of ACDs. ³ Although ACDs may streamline the compounding process, selecting an ACD to meet the needs of the institution may be challenging.

PN orders are written up and reviewed daily due to the rapidly changing clinical conditions of patients. Preparation of PN is a complex process involving the training of staff, admixing of multiple components, and complex calculations. ⁴ Our institution adopted a computerized physician order entry system for PN, with limited alerts and clinical decision support. The order was transcribed by the pharmacist and sent to the sterile compounding unit for preparation and accuracy-checking. Delay in the preparation of a PN hinders optimal patient care. Errors have been reported due to prescribing, transcribing, and preparation of PNs. ⁵ Strategies to minimize risks associated with high alert medication should be adopted.^6,7 Bonnabry et al showed a significant risk reduction to improve safety in parenteral nutrition production by redesigning the process. The most significant improvement was the elimination of fax transmission and re-transcription, and the adoption of an automated compounder. ⁸ Transcription errors may be reduced by ensuring that the hospital pharmacy system adequately integrates with the ACD. Despite the benefits associated with ACDs, there may be risks. In 2004, Baxter recalled their Exacta-Mix 2400 pharmacy compounding system due to an error with the alarm and inlet valves, allowing unwanted mixing of ingredients in the valve chamber. ⁹ However, their use may be advantageous in reducing preparation time, costs, and improved accuracy. ⁹

Although competition among vendors in providing automation solutions in KSA has been flourishing, the range of automated parenteral nutrition compounders available within KSA is limited. This study aims to assess whether automated parenteral nutrition compounders adequately meet the standards, technical and performance specifications at Prince Muhammed bin Abdul Aziz Hospital.

Method

As part of the plan to automate the parenteral nutrition process, the investigators sought to select an ACD based on a review of the technical and performance specifications within the perspective of the Ministry of National Guard Health Affairs in KSA. A task force was established, consisting of staff from the pharmacy and health technology management departments. A literature review was conducted in Medline (accessed via PubMed) and Google Scholar, including the search terms: automated compounding and parenteral nutrition. Studies published in the English language were considered. Abstracts were reviewed for their relevance to practice standards, safety, as well as technical and performance specifications. Full-text articles were sought. The searches were last updated in February 2020. Based on the literature review, the task force established a checklist of 47 technical and performance specifications. Standard terms and conditions required by the hospital were excluded from the checklist. The study excluded pricing, contractual obligations, pre-installation requirements, general technical terms, and conditions. In order to minimize bias during the evaluation, all communications were conducted through the hospital contracts department. Technical and performance specifications submitted by each vendor were reviewed and matched against the predefined checklist. In cases where the task force could not evaluate the specification due to insufficient information, the hospital contracts department sought more clarification from the vendor. The retrieved data was used to assess whether the ACDs met the standards described in the checklist of technical and performance specifications. ACDs were evaluated if the vendor submitted an offer to participate in the evaluation. A demo unit of the NutriMix KS-240 was evaluated at Prince Muhammed bin Abdul Aziz Hospital for 3 months, while the Exactamix EM 2400 was utilized at branches within The Ministry of National Guard Health Affairs. The protocol received approval from the King Abdullah International Medical Research Centre (reference number RM20/001/M), western region.

An example of the technical checklist is shown in Table 1. A summary of the final ACD evaluation was submitted to the hospital contracts department.

Table 1.

Comparison Table of Requirements Not Met by Vendors.

NutriMiX KS-240 (KAPSAM)	Exactamix EM 2400 (Baxter)
Availability of backup unit	Lack of ABACUS software integration with HIS
Lack of experience in KSA	Lack of clinical information integration
Response to downtime	Software customization by the vendor
	Lack of uninterrupted power supply
	Lack compensation for voltage fluctuations
	Lack of Service manual for HTM troubleshooting
	Consumables are vendor-specific, eg, PN bags

Note. KSA = Kingdom of Saudi Arabia; HIS = hospital information system; HTM = health technology management; PN = parenteral nutrition.

Results

Two offers were received, these included: Kapsam’s NutriMiX KS-240 and Baxter’s Exactamix EM 2400. It was found that 43 (93.48%) of the technical and performance specifications, required by the hospital, were met by the NutriMix KS-240, while 39 (84.78%) were met by Exactamix EM 2400. Table 1 summarizes the requirements that were not met at the time of the review.

A total of 42 relevant publications were retrieved from the literature. Searches were saved in Endnote citation manager. The 47 requirement standards were extracted from the relevant literature and tabulated (Table 2). The study types sought included guidelines, standards of practice, and safety. However, in the case that the standard was not relevant to the local practice, this standard was removed from the checklist. The final checklist thus numbered 46 after adjustment.

Table 2.

Technical Specifications of the Automated Compounder.

	Specification
1	Tabletop heavy-duty mixer unit to handle neonate and adult parenteral nutrition (up-to 4 liters of mixing compound) with 24-multi-source fluid compounder for micro and macro Intravenous solutions. The vendor to provide the latest in the market/industry.
2	System should include the following: a. Touch screen LCD monitor with integrated computer. b. Weighing scale. c. Privileges to database should be restricted to pharmacy admin users. Multi-user feature (up to 25 users) with unlimited number of new staff users. d. Barcode reader: with verification on setup and changing a component e. Printers: Label thermal direct printer and laser printer for reports and worksheets. f. Label: must include barcode and TALL man lettering (look-alike sound-alike drugs). The actual dose should be on the label. Label produced by 1 device only. Vendor to provide sample labels. g. Reports should include calibration and error warnings. h. Minimum fluid dispensing volume of 0.2 mL with at least a fluid dispensing resolution (increments) of 0.01 mL. Accuracy volume (%): 0.2-0.4 mL (±10%), 0.4-1.0 mL (±5%), 1 mL and above (±3%). The system should not allow manual entry if more than 0.2 mL be added to the bag. i. Emergency off (shut down) control should be included. j. Must include a backup device that handles power surges and uninterrupted power supply. k. External components can be mounted outside the laminar flow hood (eg, printer and power unit). All external components should be securely mounted l. Unit size must fit in the laminar flow hood (185 × 74 × 90 cm).
3	The following features should be provided: a. The software must be customized to use the standard units of measure used within the hospital (mEq, mmol, etc.) b. Order entry and reports customized as per hospital standard practice. c. System should be able to program a fixed sequence (physicochemical). d. Customizable soft and hard limits incorporated into the software. e. It should include weight based orders and weight based limits. f. Independent double-check on setup and change of additives, base solution, concentration and volume, tubing set traced to source solution, empty original source container presented to pharmacist for confirmation. g. Order entry/calculations should be verified by 2 staff. h. Equipment can be easily cleaned as per IV room procedures.
4	Alert should include: a. Air Bubble. b. Pharmacist override alert. c. Non empty final container (bag), empty source containers. d. Fault: pump position, valve position, opened pump cover, wrong inlet, and wrong connection. e. wrong ingredient and wrong port (barcode verification) f. Overdose / limit excess according to the patient type. g. Incompatibility of formulation (calcium-phosphate interaction). h. Insufficient source container, universal ingredient. i. Daily calibration need. j. Compounding error limit exceeded. k. Scale calibration and pump calibration calculation limit exceeded.
5	Unit should be SFDA/CE/UL approved. With proven experience/track record in KSA.
6	Vendor should provide any consumables that are needed to operate the system for at least 3 months of average usage (at least 75% of shelf life). Special tools or test equipment, hardware or software required in the troubleshooting, maintenance and performance testing of the equipment should be included as itemized price. Vendor shall provide itemized price for options, and accessories. Should allow for third party bags (flexibility of consumables).
7	Vendor shall provide 24 hour response time when the system is down with 5 years of full warranty including accessories.
8	Any utilities needed (water, vacuum, special electrical requirement, etc.) to operate the offered item should be clearly indicated into the offer.
9	Training: vendor should provide in-service to all end-users during arrival of item and also to be available anytime in need of in-service throughout the warranty period. The vendor should provide service training (with service manual) certified by the manufacturer for clinical engineering trouble shooting. This training is not required if and only if the item is serviced at the manufacturer (proof is required).
10	Power Requirements, 220 V at 60 Hz fully molded British standard power Plug (BS 1363/SASO 2203), line cord and strain relief. If unit uses a power adapter, it should be of same standard.
11	Vendor should commit to provide a replacement for the unit if it has to be serviced or calibrated outside the hospital during the warranty period.
12	Vendor should attach a copy of the catalog/brochure of the item, containing its technical specifications
13	The vendor to comply to ISD MNGHA standards for IT components including computers, servers (types, design, cabinets etc.), network, security policies (updates, patches, antivirus etc.) etc.
14	The vendor to have periodic patches and updates of operating systems (at least once per year). Preventative maintenance at least every 6 months. The vendor to install MNGHA standard antivirus if the system to be connected to hospital network.
15	Compatibility with Hospital Information System (BESTCare®): a. The unit must be able to integrate with Health Information System (HIS), pushing clinical information on to the unit. b. The unit reading must be transformed to HL7 observation result message. c. The transformed messages must be sent to the HIS using the HL7 Minimal Lower Layer Protocol (MLLP). d. The vendor responsibility to provide any software/hardware including license if necessary to fully integrate the system with the HIS.
16	The vendor to provide detailed specifications for any IT component included in their offer.

Note. cm = centimeter; CE = Conformité Européenne; KSA = Kingdom of Saudi Arabia; LCD = liquid crystal display; mEq = milliequivalent; mmol = millimole; mL = milliliter; MNGHA = Ministry of National Guard Health Affairs; SFDA = Saudi Food and Drug Authority; UL = Underwriter Laboratories.

Discussion

When selecting an ACD, organizations should carefully review the standards applied within their hospitals and the types of PN to be prepared. This will dictate factors such as the number of source containers that could be attached to the ACD. ACDs should have received the proper regulatory approval from the country of origin and within KSA. Staff should receive adequate onsite training, tools, and consumables during the initiation phase. Vendors should provide sufficient warranty periods with parts and accessories.

The workflow should include standard safety features such as: quality verification steps by technician and pharmacist during the setup, barcode verification, labels with TALL man lettering, and independent double-check with high alert medicines. Actual doses of the base solutions should be incorporated on the label. To prevent catastrophic dosing errors, hard and soft limits must be incorporated into the software. These limits could be customized by the vendor based on institutional policies. Pharmacists’ transcription of a physician order may be a potential source of medication errors. Minimizing transcription could be achieved with computerized physician order entry; however not all ACD software could be integrated with the hospital information system.

NutriMix KS-240 by KAPSAM lacked experience in KSA as the company was relatively new in the kingdom. KAPSAM provided a demo unit for a 3-month evaluation to gain more experience with the product. Some literature was in the Turkish language. This made it difficult for the task force to understand specifications and user experience. The vendor has provided English language support with specifications. At the point of the review, a backup unit was not available in the case of downtime. The response time to a maintenance request was 24 hours compared to BAXTER, which had a 3 hour response time.

The majority of the hospitals within the Ministry of National Guard Health Affairs have been using the ACD supplied by BAXTER. Ensuring medical equipment integration with current health care systems was a drawback with BAXTER. At the point of the review, the vendor stated that the ABACUS software, provided by BAXTER, could not integrate with the hospital information system, as it is not HL7 compliant. This was found to be a limitation that would hinder the hospital automation plan. The ABACUS software modifications should be completed by the vendor. BAXTER offers users the option to program the software to match local products available. However, in the case where a new concentration of a product is introduced, BAXTER would need to update the software, which may lead to delays. The software does not allow pharmacists to make changes as inadvertent changes could lead to error. The NutriSoft KS-240, provided by KAPSAM, allows predefined security levels for administrators in the pharmacy to make limited product changes or enter a new product. To prevent errors in dosing, it would be most appropriate to provide product concentrations to the vendor before initiating an ACD in an institution. Also, KAPSAM pushes software updates within 1 to 24 hours after reviewing the suggested changes by the pharmacy administrator. The Exactamix EM 2400 comes with an optional uninterrupted power supply while it comes standard with the NutriMix KS-240. This could be beneficial if there is a power outage or surge during PN preparation. Although BAXTER had not supplied a service manual, it was not necessary as the vendor conducts all services and maintenance. The PN bags for admixing was vendor-specific, and only the bags specific to BAXTER could be used. With the NutriMix KS-240, consumables are not vendor-specific, and this could be more cost-effective. Both ACDs met the minimum fluid dispensing volume of 0.2 milliliter with at least a fluid dispensing resolution (increments) of 0.01 milliliter. The accuracy, precision, and error rates were within acceptable standards.

Although vendors within KSA meet the majority of specifications and requirements, a well-designed comparative review will help in selecting an ACD. Future recommendations may include integration of clinical decision support software with an ACD software package.

Conclusion

During the comparative review of compounders available in Saudi Arabia, it was found that NutriMiX KS-240 meets the technical specifications for automated compounders required by our institution. While in the era of automation, a significant deficiency among vendors was the integration with the hospital information system. Although vendors meet many of the recommended safety standards, NutriSoft KS-240 offered HL-7 integration.