Impact of BCNSP-led parenteral nutrition management on prescribing errors in the neonatal intensive care unit

Abstract

Background

Parenteral nutrition (PN) preparations are listed as high-alert medications and have a high probability of medication errors (MEs). Board-certified Nutrition Support Pharmacists (BCNSPs) can play an important role in reducing PN-associated complications by highlighting the gaps in the PN prescribing process. This study aimed to determine the impact of BCNSP-led PN review on the identification and documentation of prescribing MEs (PMEs) to optimize quality and safety in PN prescribing processes.

Methods

This QI quasi-experimental study included all neonates admitted to a level III neonatal intensive care unit (NICU) and prescribed PN. All identified and recognized PN-PMEs documented by pharmacists were evaluated in pre-and post-implementation-phases. In the pre-phase, the PN-duty pharmacist reviewed all the neonatal PN-orders while located in the pharmacy, and in the post-phase, a clinically involved BCNSP performed this task. All PN-PMEs were categorized into ten types. Predictors of PN-PMEs were analyzed through logistic regression.

Results

PN-orders were prescribed to 98 and 112 neonates in pre-and post-phases, respectively. For demographic and clinical variables, neonates were comparable. A median of 12 (range = 9 − 19) vs. 15 (range = 12–22) PN-orders/day were reviewed in pre-and post-phases. Documented PN-PMEs for all PN orders were significantly higher in the post-phase (212/2577, 8.23%) compared with the pre-phase (25/2577, 0.97%; p < 0.001). “Wrong Dose/Calculation” was the most reported category (88/2577, 3.41%). “Wrong concentration range” was the second highest (29/1309, 2.22%) and all were reported only in post-phase. Additional errors involved infusion, compatibility, and osmolarity deviations, renal/hepatic dose adjustment errors and stability errors. Most of them were only identified in the post-phase. Post-phase outcomes showed clinically meaningful reductions in metabolic derangements, NICU stay, and mortality (p < 0.001).

Conclusion

The engagement of BCNSPs in the clinical neonatal setting for PN management may improve the safety and efficacy of PN-therapy.

Background

Critically ill neonates, especially small and sick preterm neonates, are at a high risk of medication error with compounded medications that are usually commercially unavailable and not licensed for this vulnerable population [1]. PN admixtures are listed as a class of high-alert medications (HAMs) by the Institute for Safe Medication Practices (ISMP) [2]. PN formulations require correct calculations for calories, fluid, macro- and micronutrients, total osmolarity, and adherence to compatibility limits. The large number of individual calculations accounts for a higher probability of medication errors (MEs) [3, 4].

MEs may happen at multiple stages within the medication use process, including prescribing, transcribing, preparing, dispensing, and administering, but most frequently during prescribing [5, 6]. Published safe practices guidelines assist in these processes [7–9]. A computerized prescriber order entry (CPOE) system is designed to reduce prescribing MEs (PMEs) [6, 10, 11]. Standardized neonatal PN admixtures are another way to minimize PN-related PMEs. Still, they are not always suitable for all critically ill neonates and need customization and further appropriateness checks for suitability according to individual patient conditions [12].

Several guidelines and best practices working models for the institutions have highlighted the significance and recommended the involvement of Nutrition Support Pharmacists in PN processes. These include (a) Standardized Competencies for PN Order Review and preparation, including compounding [13], (b) Best Practice Guidelines for Safe PN Practices [14], (c) American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) Standards of Practice for Nutrition Support Pharmacist [15].

In the present study, the BCNSP-led PN management intervention was developed in alignment with these established standards and safe practice recommendations. Specifically, the A.S.P.E.N. Standards of Practice and neonatal PN safety guidelines informed the structured pharmacist review process, including verification of nutrient dosing, assessment of osmolarity and compatibility limits, evaluation of clinical appropriateness based on neonatal condition, and documentation of prescribing-related interventions. This guideline-driven approach formed the basis of the BCNSP’s role in identifying and preventing PN-related prescribing medication errors within the neonatal intensive care setting.

A recent neonatal study concluded that despite the presence of clinical guidelines, neonates who received PN were exposed to different types of PN treatment-related problems. The study showed that the most common type was pharmacotherapy-related, primarily due to insufficient monitoring frequency and dosing errors. They emphasized the full and effective integration of clinical pharmacists into the healthcare team, which can serve as a potential preventive strategy to lower the occurrence of PN-related medication errors [16].

Board-Certified Nutrition Support Pharmacists (BCNSPs) receive formally designed education and are trained through a constant competency assessment for their roles and responsibilities [17]. Professional societies like ASPEN outline the standards of practice for BCNSPs [15]. BCNSPs demonstrate validated competencies in nutrition support (NS) by either completing formal learning and training programs or being certified by a credentialing body, the Board of Pharmacy Specialties (BPS), USA [15], which is established as the gold standard for determining whether a pharmacist is qualified to participate at the advanced practice level.

In many developed countries, NS-pharmacy has been evolving as a special discipline in clinical pharmacy practices (CPP). BCNSPs are qualified experts in providing several valuable services linked to PN-therapy. According to the CPP environment, the scope of BCNSPs may cover but is not limited to providing direct patient care, developing and implementing the nutritional care plan, supervising the compounding and dispensing of PN admixtures, monitoring individuals on PN-therapy, educating patients, care providers, healthcare professionals, and the public, and to contributing to quality improvement NS-related projects, research activities and administrative functions [15, 18]. Pharmacists taking part in monitoring the patients on PN therapy have long been confirmed to decrease the cost of therapy, improve clinical outcomes [19, 20], and prevent PN-complications [21].

However, evidence remains limited regarding the impact of BCNSP-led interventions on neonatal PN prescribing safety and the reduction of PN-related prescribing medication errors. Since 2017, the University Hospital has had a BCNSP involved in managing PN-therapy. Hence, the present study aimed to evalaute the impact of BCNSP-led PN management on detection of PN-PMEs and neonatal PN prescribing safety in neonatal PN prescribing processes in our hospital.

Materials and methods

Study setting and population

In this single-center quality improvement (QI) quasi-experimental study, 4 months of pre-implementation data (August-December 2016) was compared with 4 months of post-implementation data (August-December 2017). The QI intervention of BCNSP involvement in NICU as a clinical pharmacist (CP) for pharmaceutical care and NS started in March 2017. The pre- and post implementation phases intentionally selected as consecutive yearly periods to minimize temporal confounding, with the BCNSP-led intervention initiated in March 2017.

This pre-post-intervention study was conducted at a 24-bed level III NICU at the study hospital, accredited by the Joint Commission International (JCI), and a tertiary care teaching hospital associated with the university. The birth rate is nearly 5,000 births/year and approximately 16% are born premature. Yearly NICU admissions are about 1300. Critically ill neonates in the NICU are prescribed PN until the achievement of targeted nutrition through EN. All the neonates who were admitted to the NICU and prescribed PN were included. Given the retrospective quality improvement quasi-experimental design of this study, an a priori sample size calculation was not performed. Instead, all consecutive neonates who received parenteral nutrition during the defined pre- and post-intervention periods were included, representing the complete eligible population available for analysis. This approach is consistent with real-world quality improvement evaluations, where the primary aim is to assess the impact of institutional practice changes on patient care and safety outcomes.

In this study, we evaluated all the PN-PMEs reported by pharmacists over the pre-and post-implementation periods. The study has been approved by the institutional ethical review committee and the study was waived for the written informed consent as data was retrieved retrospectively.

Study outcomes

The primary outcome of this study was the rate of documented PN-related prescribing medication errors (PN-PMEs) per total PN orders reviewed during the pre- and post-intervention phases. Secondary outcomes included the classification and frequency of PN-PME types across ten predefined categories, as well as neonatal clinical outcomes associated with PN therapy, including metabolic derangements, NICU length of stay, and mortality.

Nutrition protocol for preterm neonates who received PN

A typical PN admixtures consist of PN admixture consisted of dextrose, amino acids (Aminoplasmal^® 10% B. Braun), and lipid injectable emulsions (Lipofundin^® 20% B. Braun), and micronutrients including electrolytes and vitamins. Table 1 summarizes the recommended values of PN components with the advancement of neonatal age.

Table 1.

Nutrition protocol for preterm neonates who received PN

Macronutrients and fluid	Daily values
Amino acids (g/kg/day)	Day 0: 1 Day 1: 2 From day 2: maximum of 4
Dextrose (g/kg/day)	Day 0: 8–12 From day 1: maximum of 8–14 (or higher where needed) Note: The above advancement is practiced when a patient has no issue of hyperglycemia or hypoglycemia. [Neonatal blood glucose levels are the guiding parameter for increasing/decreasing the dextrose daily dose]
Lipid injectable emulsions (g/kg/day)	Day 0: 1 Day 1: 1.5–2.5 Day 2: 2–3 Day 3: 2.5–3.5 From day 4: maximum of 3
Fluid volume (ml/kg/day)	Day 1:80–100 Day 2: 110–120 Day 3:120–140 Day 4 & 5:140–160 From day 6: maximum of 160–180 to 200, depending on the neonate’s birth weight

Process of neonatal parenteral nutrition at study hospital

1. Pre-intervention phase

• i)PN prescription and calculation

In the Pre-Imp-Phase the assigned post-graduate residents of neonatology assessed the requirement for PN during daily NICU rounds with a team of attending neonatologists, followed by PN calculations for required components, considering the recommended dietary allowances (RDAs), recently reported biochemical profiles, fluid balance, and blood glucose levels of the admitted infant. Most residents have 2–4 months of rotation in the NICU and are assigned to 5–6 neonates daily starting from their first day in the NICU. PN prescription requests to the pharmacy are generated through a CPOE system.

• ii)PN review process by duty pharmacist

After prescriber order entry, the PN-orders appeared at the pharmacist’s online CPOE system screen in the clean room pharmacy, where a pharmacist on PN duty performed review of all neonatal PN orders, including verification of patient data, PN-components, and appropriate dosages matching with weight and age. The PMEs included in this study were defined as physicians’ unintended deviations from a “defined standard,” that might result in patient harm. Manufacturer information, institutional neonatal PN protocols, and internationally applicable consensus guidelines were used as the defined prescribing standards, with recommended PN component dose ranges presented in Table 1 [22–27].

For any necessary modification in the PN order, this pharmacist had to communicate with the ordering prescribers in the NICU over the phone. Afterward, the requests were sent for sterile compounding to the PN technician. The pharmacist on PN duty was not a BCNSP and had 1–3 years of postgraduate experience.

• iii)Documentation of PN-related prescribing errors

All the PN-PMEs were documented in the CPOE system by the PN-duty pharmacist as pharmacist interventions.

• iv)PN administration by the nurse

Once all the PN admixtures were received from the pharmacy, the neonatal assigned nurse checked the label details with the prescription prior to administering the PN admixture.

• 2. QI Intervention: implementation phase

In 2017, a BCNSP was assigned as a neonatal CP for pharmaceutical care including PN therapy. In the QI implementation phase, the BCNSP was involved in daily NICU rounds with faculty (mostly 9 am to 12 noon) and had a thorough review of each NICU patient every 24 h, except weekends. Following the pharmacy work rules, a pharmacist’s review is mandatory for all admixture orders such as PN before preparing it (after rounds, mostly 12 noon till 2 pm). Therefore, for neonatal PN orders at the hospital, this task was assigned to the BCNSP, a dedicated neonatal CP. During this period, the BCNSP had an opportunity to develop good working relationships with the clinical team.

• 3. Post-intervention phase

In this phase, PN orders were placed by the NICU residents but reviewed by the BCNSP on the unit for all necessary aspects including verifying patient data, PN composition, appropriateness of dosages per age, body weight, and clinical condition, adherence to osmolarity limits and maximum electrolyte concentrations for administering through the peripheral route. The BCNSP reviewed the orders for completeness with respect to incompatibility and stability of the final PN admixture. The BCNSP also considered all medications administered to neonates, for any drug-drug or drug-PN interaction, and the total daily intake of nutrients, fluid, minerals, and electrolytes through PN and EN. During the post-phase, all the PN-PMEs were documented daily in the CPOE system or patient file by the clinically assigned BCNSP. On weekends and public holidays, the PN duty pharmacist logged the pharmacist interventions in CPOE for all intervened PN orders to ensure continuity of reporting.

Data collection

During both phases, the data were retrieved for PN-PMEs through the number of pharmacist interventions logged in the CPOE system and patient files during hospitalization. PN prescription data were extracted directly from the CPOE system, including ordered amounts of macronutrients (dextrose, amino acids, and lipid emulsions) expressed as weight-based daily doses (g/kg/day), as well as electrolytes and micronutrients recorded as mEq/kg/day. Additional extracted fields included total prescribed fluid volume (mL/kg/day), route of administration (central vs. peripheral), calculated osmolarity, and pharmacist intervention notes documenting PN-related prescribing medication errors. Data for neonatal demographic and clinical characteristics were retrieved, including date of birth and duration of hospitalization, gestational age, gender, and birth weight. Detailed errors according to the PN component such as amino acids (AAs), fluid, etc. were also gathered through the pharmacist intervention data and files. In addition, data was retrieved for PN indications and PN-related metabolic complications. Metabolic complications were identified through routine laboratory monitoring documented in the electronic medical record, including serum electrolytes, glucose and liver function tests (LFTs) and blood urea nitrogen (BUN). Data for intravenous fluids (other than PN), contrast agents, vaccines, oxygen therapy, blood products, and topical applications were not included in the analysis. The data were extracted retrospectively and a team of pharmacists and physicians from the list of researchers performed the analysis.

Operational definitions

Medication errors were defined as per the US National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP) definition [28]. All PN-related prescribing medication errors (PN-PMEs) that resulted in a change in the PN order after discussion with the ordering physician were documented as PN-related PMEs. These errors were classified into ten predefined categories, including wrong patient/data or PN not needed, wrong dose/calculation, renal/hepatic dose adjustment errors, wrong rate of infusion or method, wrong route, wrong indication, missing or late PN order entry, compatibility and stability errors, unsafe concentration range, and osmolarity errors, as summarized in Table 2.

Table 2.

Types of PN-related prescribing errors* and their description

Types	Description of PN-related prescribing errors
Wrong patient/Patient data/Patient not needed PN	➢Wrong patient name, date of birth, or weight ➢PN was prescribed for the patient who did not need it on that particular day (as decided during the multidisciplinary team’s daily round, either to discontinue PN or not to start) ➢Patients were prescribed a PN admixture based on another patient’s lab values
Wrong Dose/Calculation	➢Prescription of the wrong component in PN ➢This included a wrong choice of drug concentration (e.g., 5% glucose instead of 50% glucose) ➢The patient was prescribed the wrong dose or formulation of PN admixture [Drug dosage is specified daily and adjusted to body weight] ➢The patient was prescribed the wrong dose of any component of the parenteral nutrition (PN) admixture
Renal/Hepatic Dose Adjustments Errors	➢Prescribed PN without adjusting the components adjustments according to renal or hepatic functioning
Wrong rate of infusion/Method	➢The PN admixture was prescribed via the wrong method (received intermittently instead of continuously) ➢The PN admixture was prescribed with the wrong rate of infusion
Wrong Route	➢The wrong route of administration was selected while prescribing PN order
Wrong Indication	➢Prescription of a drug without indication, or omission of a drug even though the drug/PN component was indicated
Missing to enter PN order/Late entry.	➢Missed entering PN order for any patient ➢Entered late PN order in CPOE
Compatibility and stability Errors	➢Prescription of PN admixture that may be chemically incompatible [e.g., The lack of amino acids in PN is regarded as a potential cause of precipitation of calcium phosphate in admixture simultaneously containing calcium gluconate and sodium glycerophosphate, Equally, the lack of amino acids in PN exposes water-soluble vitamins to an increased degradation through trace elements] ➢Y-site incompatibility with our medications/products
Unsafe Concentration Range	➢The wrong concentration range was prescribed in PN, especially dextrose above 12.5% given via peripheral line and a concentration range of calcium gluconate above 0.4% in PN that is applied via a peripherally inserted venous catheter
Osmolarity Errors	➢The osmolarity of a PN that is applied over a peripherally inserted venous catheter should be lower than 750 mosmol/l. Osmolarity from 750 mosmol/l to 800 mosmol/l is accepted. An error is defined as an osmolarity above 800 mosmol/l

* The National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP) index was used to categorize the outcome of medication error

PN-associated metabolic derangements were defined as clinically significant abnormalities in monitored laboratory parameters (e.g. sodium, potassium, LFTs and BUN) requiring medical or PN prescription adjustment, according to institutional neonatal reference ranges [20]. Serum potassium level of 3.5–6.0 mmol/L was considered normal, and hyponatremia was defined as sodium serum levels of < 135 mEq/L. Serum glucose greater than 150 mg/dl was considered as hyperglycemia [29].

Statistical analysis

Statistical analysis was performed using STATA version 15 (STATA Corp, Texas). Comparisons between Pre-Imp-Phase and Post-Imp-Phase were completed by applying the student t-test or Mann-Whitney test. The Chi-square (χ2) test was used to assess the association between categorical variables. The Fisher’s exact test was applied to establish the significance where the expected frequencies were < 5. To assess the association between BCNSP involvement and PME types, logistic regression was utilized to determine the odds ratio at a 95% confidence interval. Absolute and relative frequencies of errors were calculated. The p-value < 0.05 was considered statistically significant. Logistic regression assumptions were assessed by ensuring binary outcome classification and absence of multicollinearity among predictors. PN orders were analyzed at the order level as pharmacist interventions, consistent with the study objective of evaluating prescribing deviations per PN prescription. The dataset was complete for all study variables; therefore, no missing data handling or imputation was necessary.

Results

Total number of documented medication errors

A total of 201 and 222 infants were admitted to the NICU during the Pre-Imp-Phase and Post-Imp-Phase respectively. Of these, 98 neonates in the Pre-Imp-Phase and 112 neonates in the Post-Imp-Phase were prescribed PN. In both study phases, most neonates received PN due to prematurity (e.g., GA < 32 weeks and/or birth weight < 1800 g). Other major PN indications for preterm babies were that EN was stopped for > 24 h and it was unlikely to restart and make sufficient progress with EN within a further 48 h. The gestational age, birth weight, and length of hospital stay of patients on PN were comparable in both phases; most were male (Table 3). During the pre-phase 12 PN orders/day (range 9 − 19) were reviewed by PN duty-assigned pharmacists, who documented 25 PN-PMEs while processing 1268 PN-orders (1.97%). However, during the post-phase 15 PN-orders/day (range12–22) were reviewed by the BCNSP, who identified and documented 212 PN-PMEs, a significantly higher (p < 0.001) number while processing 1309 PN-orders (16.19%). Two non-prescribing pharmacist interventions recorded during the post-intervention phase were excluded from analysis as they did not meet the definition of PN-related prescribing medication errors. Most of the analyzed PN orders were prescribed by the residents of years 1 and 2 (Table 3).

Table 3.

General characteristics of the neonates and PN prescribing errors

Number of patients admitted to NICU*	pre-phase	post-phase	Total	p-value **
	201 (47.52)	222 (52.48)	423 (100.0)	0.573
Number of patients prescribed PN order*	98 (48.76)	112 (50.45)	210 (49.65)	0.197
Gestational age (weeks)	31.7 ± 3.1	31.2 ± 1.2	31.3 ± 4.2	0.687
Birth Weight (g)	1,110.8 ± 550.6	1018 ± 301.6	1048 ± 379.9	0.489
Male	59 (60.2)	65 (58.0)	124 (59.0)	0.592
PN orders/day a	12 (9–19)	15 (12–22)	13 (10–22)	0.315
Total number of PN orders *	1268 (49.20)	1309 (50.80)	2577 (100.0)	0.061
Total PN via the peripheral route *	304 (23.97)	378 (28.87)	682 (26.46)	0.084
Documented PN-related prescribing errors/day a	1 (0–2)	7 (4–9)	3.5 (3–4)	< 0.001
Total number of documented PN-related prescribing errors *	25 (1.97)	212 (16.19)	237 (9.19)	< 0.001
PN Prescribed by a medical provider
Resident year- 1	372 (29.34)	398 (30.40)	770 (29.88)	0.421
Resident year- 2	591 (46.61)	586 (44.77)	1177 (45.67)	0.865
Resident year- 3	185 (14.59)	192 (14.67)	377 (14.63)	0.531
Resident year- 4	120 (9.46)	133 (10.16)	253 (9.82)	0.802
Indications of PN * $	pre-phase n = 98	post-phase n = 112	Total n = 210	p-value
For preterm babies born before 31 + 0 weeks	36 (36.7)	45 (40.2)	81 (38.6)	0.468
For preterm babies born at or after 31 + 0 weeks and sufficient progress not made with EN in the first 72 h of life.	15 (15.3)	20 (17.9)	35 (16.7)	0.719
PN for preterm and term babies who could not establish sufficient EN • a congenital gut disorder • a critical illness such as sepsis.	13 (13.3) 11 (11.2)	11 (9.8) 13 (11.6)	24 (11.4) 24 (11.4)	0.204 0.926
For preterm babies, EN was stopped for > 24 h and it was unlikely to restart and make sufficient progress with EN within a further 48 h.	18 (18.4)	20 (17.9)	38 (18.1)	0.890
For term babies, EN was stopped for > 48 h and it was unlikely to restart and make sufficient progress with EN within a further 48–72 h	5 (5.1)	3 (2.7)	8 (3.8)	0.351
Patient safety and LOS *
PN- Associated Metabolic derangements*
Overall observed Metabolic derangements	164 (72.57)	62 (27.43)	226 (100)	< 0.001
Deranged electrolytes	54 (32.93)	24 (38.71)	78 (34.51)	0.035
Deranged LFTs/raised DB	50 (30.49)	19 (30.64)	69 (30.53)	0.081
Hyperglycemia	36 (21.96)	9 (14.52)	45 (19.91)	< 0.001
Raised BUN	24 (14.63)	10 (16.13)	34 (15.04)	0.030
Length of stay and mortality*
Discharged within 7 days*	2 (2.1)	4 (3.6)	6 (2.9)	0.12
Discharged within 8–14 days*	21 (21.4)	32 (28.6)	53 (25.2)	< 0.001
Expired within 7 days*	5 (5.1)	3 (2.7)	8 (3.8)	0.11
Expired within 8–14 days*	9 (9.2)	2 (1.8)	11(5.2)
Overall mortality*	21 (21.4)	11 (9.8)	33 (15.7)	< 0.001
Total days on PN a	9.0 (6.0–26.0)	5.0 (4.0–14.0)	7.5 (4.0–26)	< 0.001
LOS in NICU (days) a	10.0 (6.0–21.0)	8.0 (5.0–18.0)	9.5 (5.0–21)	< 0.001
LOS in hospital (days) a	11.5 (6.0–28.0)	10.5 (7.0–27.5.0.5)	12.5 (6.0–27.5.0.5)	0.588

Normally distributed data, expressed as mean ± standard deviation

^a non-normally distributed data expressed as median (25–75 quartiles);

* Data presented as n(%)

BUN= blood urea nitrogen; DB=Direct bilirubin; GIT= Gastrointestinal tract; LFTs=liver function tests; LOS= length of stay; NEC = necrotizing enterocolitis

^$ PN Indications are defined for data analysis, according to National Institute for Health and Care Excellence(NICE) Neonatal PN Guidelines 2020

Neonates in the post-phase had significantly lower PN-associated metabolic derangements during and after PN therapy, length of NICU stay (p < 0.001), and mortality. A significantly higher number of neonates were discharged within 8–14 days of life (p < 0.001) (Table 3).

PN-related prescribing errors categories

The PN-PMEs were classified into ten error-type categories (Table 4). The most frequent category was “Wrong Dose/Calculation,” accounting for 88 cases (37.1% of all PN-PMEs and 3.4% of all PN orders). These errors are clinically significant in neonates because inappropriate macronutrient or electrolyte dosing may increase the risk of metabolic instability and PN-related complications. Amino acids (n = 35), sodium (n = 17), and potassium (n = 8) were the most commonly involved components.

Table 4.

Frequency of documented PN-related prescribing errors by type in the pre- and post-implementation phases

	Pre-phase		Post-phase		OR
	n *	Documented Errors n (%) **	n *	Documented Errors n (%) **	(95%CI)	p-value
Wrong patient/Patient data	1268	0 (0.00)	1309	10 (0.76)	-
Wrong Dose/Calculation		12 (0.95)		76 (5.81)	3.84 (3.26–7.05)	< 0.001
Renal/Hepatic Dose Adjustments Errors		0		23 (1.76)	-	-
Wrong rate of infusion/Method		6 (0.47)		22 (1.69)	1.32 (1.09–2.07)	< 0.001
Wrong Route		4 (0.32)		14 (1.07)	1.92 (1.09–3.52)	< 0.001
Wrong Indication		3 (0.24)		17 (1.30)	1.27 (1.11–3.05)	< 0.001
Missing to enter PN order		0		9 (0.69)	-	-
Compatibility and stability Errors		0		8 (0.62)	-	-
Wrong Concentration Range		0		29 (2.22)	-	-
Osmolarity Errors		0		4 (0.31)	-	-
Total		25 (1.97)		212 (16.20)	7.24 (3.06–9.53)	< 0.001

* n: total number of prescribed PN orders;

** n (%): absolute frequency (relative frequency) of the errors;

*** p-value < 0.05 indicating statistically significant differences among Pre and Post phases (logistic regression)

95%CI: 95% confidence interval

Table 5.

Frequency of documented PN-related prescribing errors according to the PN components in the post-implementation phase

PN Component	Times involved in prescribing errors n (%)
Amino Acids	41 (22.4)
Dextrose	35 (19.1)
Lipid	31 (16.9)
vitamins	12 (6.6)
Electrolytes	33 (18.0)
Trace Elements	9 (4.9)
Fluid	22 (12.02)
Total	183 (100)

The second most frequent PME type was “Wrong concentration range,” with 29 cases (12.2% of PN-PMEs; 1.13% of PN orders), all reported in the post-intervention phase. Such errors may lead to excessive osmolarity and increase the risk of peripheral vein complications. “Wrong rate of infusion/method” errors represented 28 cases (11.8% of PN-PMEs), while “Renal/Hepatic dose adjustment” errors accounted for 23 cases (9.7% of PN-PMEs), mainly involving lipid injectable emulsions. “Wrong indication” errors were also observed in 20 cases (8.4% of PN-PMEs), including omission of amino acids or multivitamins (Table 4).

“Wrong route” was identified in 18 cases (7.6%), primarily when central PN was prescribed through a peripheral line. Compatibility and stability issues were less frequent (8 cases, 3.4%), including documented Y-site incompatibilities. Errors related to wrong patient/patient data occurred in 10 cases (4.2%), while osmolarity-related errors were identified in 4 cases (1.7%), resulting in excessive osmolarity intended for peripheral administration. In phase-II, nine times BCNSP asked the prescriber to enter the missed PN order in CPOE.

Discussion

By introducing a BCNSP to the neonatal ICU and involving the pharmacist in direct patient care, we observed a statistically significant increase in PME detection following BCNSP implementation (p < 0.001), along with a significant reduction in PN-associated metabolic derangements in the post-intervention phase (p < 0.001). The BCNSP’s unique combination of medication and nutrition knowledge and qualifications was key to the success of this initiative.

Existing literature on pharmacist involvement and PN safety in neonates is limited but suggests meaningful benefits in error prevention and quality improvement. A broader review of neonatal parenteral nutrition practices highlights the importance of pharmacist participation in PN order verification and clinical decision-making, although high-quality outcome data remain scarce [30]. In addition, observational studies of neonatal PN have documented frequent drug-related and treatment-related problems, emphasizing the high-risk nature of PN therapy in this vulnerable population and the potential value of specialized pharmacist expertise in optimizing safety [31]. Within this context, Hermanspann and colleagues demonstrated that clinical pharmacist cross-checking can identify prescribing errors and recommended its routine inclusion to prevent adverse events [32]. Similarly, our study focused on PN-related prescribing medication errors and found a higher documented PME rate (16.2%), which likely reflects enhanced interception and systematic documentation of near-miss prescribing deviations through BCNSP involvement rather than an actual increase in prescribing errors. Importantly, these events were detected and corrected before PN compounding and administration. Routine dietitian involvement in PN order review was not reported in Hermanspann et al. and was also not part of our PN verification workflow.

Few neonatal studies have reported the impact of different study interventions on PN-PMEs. Brown and colleagues [6], reported a reduction in neonatal PN-PMEs from 14 to 7% by establishing a computer-generated PN form. In another neonatal study by Lehmann and colleagues [33], a 6.6% reduction in PN-PMEs was reported after introducing the CPOE system. Hilmas and colleagues [34], reported that a clinical pharmacist in a children’s hospital intervened in > 5% of PN prescriptions. Similarly, in the present study, the involvement of BCNSP in PN order review and the physical presence of BCNSP in the NICU resulted in higher documentation of PN-PMEs and avoided adverse drug events.

The classification of MEs was different in previous studies. Brown and colleagues [6], included calcium-phosphate solubility, total volume, and osmolarity errors in the list of PN-PMEs. On the other hand, MacKay and colleagues [35], defined compatibility MEs as PN-administration errors. In contrast to previous studies [32], our study segregated PN-PMEs into ten predefined categories, which can further help to differentiate the knowledge gap and deficiencies in the system. Based on these results, strategies can be developed to improve the PN knowledge of prescribers and pharmacists and develop system-based solutions to prevent PN-PMEs.

The present study reports that pharmacists were generally executing traditional responsibilities to formulate and dispense PN bags in the pre-phase. Similar general practices of PN-duty pharmacists are reported in another hospital study where the pharmacist was primarily involved in technical responsibilities such as PN-order processing, checking dose/weight [mentioned in the computer system], supervising PN compounding technicians, labeling, and dispensing [36]. Most pharmacists working out of the dispensary have very little direct interaction with patients and prescribers. Being on the unit and clinically involved in the patient’s overall care makes the pharmacists’ involvement in PN more valuable. Clinically integrated BCNSP-led PN-management has been shown to improve the identification and documentation of PN-associated PMEs [37, 38]. Our study reflects the same benefits of involving BCNSP in PN management, which are highlighted in several guidelines and recommended best-practice competency frameworks for the institutions [13–15].

In this study, the CPOE system and pharmacists were involved in both phases, but involving a BCNSP as a CP helped the system to highlight the gaps in PN prescribing process. The BCNSP’s involvement in PN prescription review and documenting the prescription errors to correct them before formulation and dispensing may protect this vulnerable population from significant adverse effects along with achieving the optimal neonatal daily growth needs.

The present study has several limitations. It was a single-center study conducted over a limited timeframe, which may affect the generalizability of the findings. Although quasi-experimental designs are inherently subject to potential confounding, the same CPOE system and comparable prescriber experience were maintained across the pre- and post-intervention phases. Nonetheless, unmeasured factors and temporal variations may still have influenced prescribing practices. Because multiple PN orders could be generated for the same neonate, some within-patient correlation may exist. Given the retrospective quality improvement design, a formal a priori sample size calculation was not performed; instead, all consecutive neonates receiving PN during the study periods were included, representing the complete eligible population available for evaluation.

This study has several strengths, including the evaluation of a substantial number of neonatal PN orders and the structured involvement of a BCNSP in daily clinical review and pharmacist–prescriber communication. The quasi-experimental design allowed assessment of practice change following BCNSP implementation. To our knowledge, this is the first study in our country to examine the impact of BCNSP-led PN management on identifying PN-related prescribing medication errors in the NICU, providing valuable evidence to support safer neonatal PN practices.

Conclusion

The engagement of the BCNSP in neonatal PN management resulted in significantly higher detection of prescribing deviations per PN order, reflecting improved interception of near-miss errors before compounding and administration. These findings highlight the important role of BCNSPs in enhancing the safety and quality of PN therapy for critically ill neonates. With their specialized competencies in nutrition support, BCNSPs are well-positioned to support safe PN prescribing through comprehensive order review, interdisciplinary collaboration, and prevention of PN-related medication errors in this high-risk population.

Future recommendations

The emphasis of future studies shall be on investigating the strategies to make PN practices safe and effective by utilizing the knowledge and skills of BCNSPs, including educational interventions for pharmacists, prescribers, and nurses. Directing BCNSPs for PN prescribing rights can be an appropriate strategy to improve the execution of collaborative patient care. Furthermore, future studies are also warranted to estimate the influence of pharmaceutical care services on the clinical and economic outcomes once these services are established systemically by BCNSPs in the country.

Abbreviations

AAs

Amino acids

ASPEN

American society for parenteral and enteral nutrition

BCNSP

Board-certified nutrition support pharmacist

CP

Clinical pharmacist

CPOE

Computerized prescriber order entry

EN

Enteral nutrition

HAM

High-alert medications

ISMP

The institute for safe medication practices

JCI

Joint commission international

NCC MERP

National coordinating council for medication error reporting &prevention

NICU

Neonatal intensive care unit

NS

Nutritional support

PEMs

Prescribing medication errors

PN

Parenteral nutrition

PN-PMEs

PN-associated PMEs

RDAs

Recommended dietary allowances

Author contributions

GA has substantial contributions to conceptualizing and designing the project. GA, SA, AUR, IH, AA, HMAY, and MK were involved in data collection and data entry. GA wrote the manuscript. KH and MSS were involved in supervising the project, interpreting data, and substantively revising the manuscript. GA, SA, AUR, IH, AA, HMAY, and MK analyzed the data and completed the analysis writeup. ZB and RM carried out data analysis, wrote the initial revised analysis writeup, and contributed significantly to the revision phase. AS and HMAY wrote the revised manuscript draft and contributed significantly to the revision phase. GA, KH, and MSS completed the final draft and proofreading. All authors were equally involved in manuscript writing and critical revising. All authors reviewed the manuscript.

Funding

This work is not funded.

Data availability

All data generated or analyzed during this study are included in this published article. The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The research study was performed in accordance with all the relevant guidelines and regulations of the Declaration of Helsinki. Before performing this study, ethical approval was taken from the institutional ethical committee of Aga Khan University Hospital (ERC # 2020-5725-15211) and informed consent from a parent or guardian for participants was waived by ERC as all the data was collected retrospectively.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.