3.3.5 Economic benefits of PN

CRITICAL ILLNESS

• The PEPaNIC trial demonstrated that delaying PN during the first week of critical illness in children is not only clinically superior but also cost-effective. In a cost analysis involving Belgian and Dutch patients, late PN reduced direct medical costs by 21% (€7,180 less per patient) compared to early PN. The primary driver of cost savings was reduced intensive care hospitalisation costs, while the cost of PN itself accounted for only a small portion of the savings. New infections, which were significantly lower with late PN, contributed substantially to overall costs, with 14% of patients incurring 41% of total expenses. Overall, late PN proved both more effective and less costly, making it the superior strategy.³⁴⁴
• Doig and colleagues (2013) conducted a cost-minimisation analysis from the perspective of the US acute care hospital system to estimate the cost implications of providing early PN (within 24 hours of ICU admission) to patients with short-term relative indications to early ETF. Clinical outcomes and measures of resource consumption were taken from a multicentre clinical trial involving 1,363 patients, combined with cost distributions obtained from the literature. The analysis showed that early PN significantly and meaningfully reduced the total costs of acute hospital care by US$3,150 per patient (95% CI 1,314–4,990). (Mean costs of ICU care were $58,924 [95% CI 57, 631–60,239) with standard care and $55,772 [95% CI 54,484–57,082] with early PN.) All sensitivity analyses demonstrated significant cost savings with early PN, including use of European cost data.¹⁴⁶
• Pradelli and colleagues used discrete event simulation and a deterministic simulation model to evaluate the cost-effectiveness of SPN (ETF + PN) administered on days 4–8 of ICU admission compared with continued ETF in patients who did not achieve ≥60% of their targeted energy intake by day 3. Total hospitalisation costs were estimated at 112,338 CHF per patient receiving EN and 108,999 CHF per patient receiving SPN, resulting in an estimated net cost reduction with SPN of 3,339 CHF per patient. Each 1,000 kcal decrease in cumulative energy deficit with SPN was associated with a 10% reduction in the risk of nosocomial infection (OR 0.90; 95% CI 0.83–0.99; p < 0.05). The 5.3% absolute reduction in nosocomial infection with SPN yielded a number needed to treat to avoid one infection of 19, with a saving of 63,048 CHF per infection avoided. The cost of the intervention was therefore more than offset by the cost saving realised through the reduction in nosocomial infection.¹⁴⁵
• The Swiss SPN study showed that in critically ill adults who fail to meet energy targets with EN alone, the addition of SPN from days 4–8 significantly reduces energy deficit and lowers the risk of nosocomial infections. A cost-effectiveness analysis revealed that each 1000 kcal reduction in energy deficit was linked to a 10% drop in infection risk. Importantly, the strategy was cost-saving, with an estimated reduction of 63,048 CHF per avoided infection, highlighting SPN as an economically favourable approach in this patient population.¹⁴⁵
• A targeted literature review and exploratory cost-utility analysis evaluated the potential cost-effectiveness of SPN compared to EN alone in cancer patients, particularly when EN fails to meet nutritional requirements. The review linked SPN to improvements in key nutritional markers—BMI, fat-free mass, phase angle (PhA), and prealbumin—two of which (BMI and PhA) are strong survival predictors. Indirect estimates suggested that SPN could reduce mortality risk, with hazard ratios ranging from 0.80 to 0.99. In a model focusing on patients with stage IV inoperable pancreatic cancer, the incremental cost-effectiveness ratio of SPN varied between £41,350 and £91,501 depending on how service costs were accounted for. While direct clinical evidence is limited, findings indicate that SPN may offer significant clinical and quality of life benefits, with greater cost-effectiveness likely in less advanced disease stages or other cancer types.³⁴⁵
• Not all studies have reported cost savings with early use of PN in the ICU. Vanderheyden and colleagues reported the cost analysis from the 2007–2010 EPaNIC trial¹³⁶ Early PN (n = 2,312) was associated with mean costs of €17,973 (SD €18,965), compared with €16,863 (SD €18,190) for late PN (n = 2,328), a difference of €1,110. However, the cost increment with early PN was only €94 per patient in those who did not develop an infection and who did not require prolonged ICU support, who comprised 70.1% of the total population. The EPaNIC trial’s cost analysis showed that initiating PN early in critically ill ICU patients led to higher healthcare costs without clinical benefit. Early-PN increased pharmacy-related expenses—especially for PN itself and anti-infective agents—by an average of €1,210 per patient. Patients who developed infections or had prolonged ICU stays contributed disproportionately to total costs. Given the lack of improved outcomes and increased expenditures, delaying PN (Late-PN) is a more cost-effective strategy in this population.³⁴⁶ In a cost-effectiveness analysis compared PN to EN in critically ill adults using data from a large UK multicentre trial. At both 90 days and one year, quality of life and survival outcomes were similar between groups, but PN was associated with higher costs and a negative incremental net benefit (INB) at one year (−£1,320). When projected over a lifetime, the INB for PN turned slightly positive (£440), though with substantial uncertainty. Overall, PN is unlikely to be cost-effective compared to EN in critically ill adults, especially in the short term. ³⁴⁷ Furthermore, as discussed in Section 3.3.1.1, the trial on which this analysis is based had several important methodological limitations that are likely to have biased the cost analysis in favour of late PN. For instance, patients in this study remained in the ICU for only a short time (average, 3 4 days in >70% patients), which may not have been long enough to demonstrate a benefit with early PN. Also, 75% of patients had a normal or slightly higher than normal BMI,¹⁴⁷ suggesting that this population may not be representative of the population typically indicated for early PN in the ICU setting. Clinical outcomes in patients who are not malnourished are less likely to be altered by the addition of PN.

• Shields and colleagues reported on the long-term cost-effectiveness of in-hospital TPN over 10 years (1983–1993) in 162 patients with acute GI failure (4,997 patient-days; 192 central venous catheters). In patients with non-malignant disease, fed for >21 days (mean 50 days), 10-year survival was 74%, at a cost of £4,723 per year of life saved; in patients with malignant disease, 5-year survival was 27%, at a cost of £8,351 per year of life saved. On the assumption that TPN was lifesaving in patients who need long-term TPN, these costs were considered to compare favourably with other technologies such as dialysis for ARF, which has an in-hospital mortality rate of 50% (even with dialysis). Treatment by an expert team improved patient selection and complications rates, and reduced costs.³⁴⁸ Although this study was published some years ago, and therefore may not reflect current practice, PN is still likely to compare favourably with other supportive technologies used in the ICU, as in-hospital mortality for ARF remains high (up to 60%) and is associated with significant costs.³⁴⁹

PN can be delivered in different ways, ranging from bespoke pharmacy-compounded PN to commercial premixed multichambered bags

• Alfonso and colleagues (2016) conducted a systematic literature review of studies comparing MCB and pharmacy-compounded PN from January 1990 to November 2014; 18 published studies (mostly retrospective) met the inclusion criteria. Ten studies (including one prospective randomised trial and multiple retrospective analyses) reported a lower risk of BSI with MCBs compared with other delivery systems. Sixteen studies reported ergonomic and/ or economic outcomes; most reported a potential cost benefit with MCB, with consistent reports of reduced time and labour compared with other systems. The largest cost benefit was seen in studies that evaluated total hospitalisation costs (e.g., costs relating to infectious complications and length of stay in the hospital or ICU). The authors noted that methodological factors limited the quality of the evidence.³⁵⁰ Furthermore, none of the included studies evaluated errors associated with the PN process, which have been shown to be reduced for MCB PN compared with compounded PN preparation.³⁵¹
• In a prospective, observational, cost-accounting study conducted in 10 Spanish hospitals, MCBs were compared with hospital-compounded bags (COBs). MCBs demonstrated greater cost-effectiveness, saving an average of $5.71 per bag ($62.11 vs. $67.54), reducing preparation time by 38 minutes, and significantly lowering preparation errors (1.0% vs. 5.0%). These findings suggest MCBs may reduce the economic and clinical burden associated with PN.³⁵²

HPN

• An observational, retrospective study aimed to evaluate the direct healthcare and non-healthcare costs associated with the HPN program at Gregorio Marañón University Hospital in Madrid, Spain, over the period from January 2014 to October 2015. The analysis included 32 adult patients with chronic intestinal failure. Data were collected from medical records, dispensary services, and the hospital’s financial services. The total direct healthcare and non-healthcare costs per patient were €13,363.53, or €124.02 per patient per day. Direct healthcare costs made up 98.32% of the total costs, with HPN provision accounting for the largest portion at 74.25%. The remaining 1.68% of costs were attributed to non-healthcare expenses, primarily transportation. The study underscores that direct healthcare costs, particularly those associated with HPN provision, are the dominant component of overall HPN expenditures.³⁵³
• A study evaluated the cost-effectiveness of supplemental home parenteral nutrition (sHPN) in patients with incurable gastrointestinal cancer in China. Clinical data were taken from a randomised controlled trial and analysed using a Markov model. The incremental cost-effectiveness ratio (ICER) for sHPN was $24,289.17, with an incremental cost of $2,051.18. and an incremental QALY of 0.0844 compared to non-sHPN care. Sensitivity analyses confirmed that sHPN remained cost-effective even with variations in utility values. The results suggest that sHPN is a cost-effective option for this patient population in China, supporting its clinical application and offering valuable insights for decision-making and pricing.³⁵⁴
• Naghibi and colleagues conducted a systematic review and meta-analysis of survival, QOL, and cost-effectiveness of HPN in patients with inoperable bowel obstruction, based on 12 studies involving 437 patients. In the base-case analysis, the incremental cost for HPN over no nutritional treatment was £22,197, yielding an incremental cost-effectiveness ratio (ICER) of £176,587 per quality-adjusted life-year (QALY) gained. However, this ICER was highly sensitive to the utility values in the treatment group (−30% to 75%), and to the cost of PN (±26%) and survival of the treatment and non-treatment groups (−24% to +10% and −20% to +7%, respectively), although ICERs remained above £123,000 per QALY gained in all sensitivity analyses. The daily cost for HPN was £240. The authors comment that whilst the costs of HPN are high, they are comparable to those for the “less controversial and accepted practice of HPN for benign disease”. Performance status and predicted survival are key factors in deciding on HPN for a patient with malignant bowel obstruction. The authors also note that although cost-effectiveness analysis is the dominant method used in health economics to value health technologies, experts suggest that the willingness to pay threshold (i.e., the maximum amount of money that a healthcare system is prepared to give up to ensure that a health technology is implemented³⁵⁵) may be more relevant for the evaluation of interventions in palliative care, as it is typically higher (e.g., up to £70,000 per QALY in the UK) than the ICER threshold.³⁵⁶ It is also important to note that some health technology assessment agencies (e.g., the National Institute for Health and Care Excellence [NICE] in the UK) apply additional criteria to end-of-life treatments.³⁵⁷ Treatments that meet these criteria may be recommended at a higher ICER threshold.

• HPN has a key role in shortening the hospital stay for patients who are ready to be discharged but who require IV nutrition,³⁵⁸ which is likely to realise considerable cost savings for many healthcare systems.
• An economic analysis by Marshall and colleagues found that HPN was significantly cost saving compared with hospital PN in Canada. In this study, the cost of home and hospital PN was compared through detailed review of the medical records for all patients managed by an HPN programme between 1996 and 2001 whose PN was initiated in hospital (n = 29). Direct medical costs were estimated for the 2 weeks before hospital discharge and for the first month after discharge home. Common indications were malignancy, IBD, and intestinal ischemia (n = 12, 6, and 4, respectively). Mean daily costs were higher in in the last week of hospitalisation than in the first month of discharge ($567 vs $405; p < 0.0001). Acute care accounted for <10% of overall costs on HPN. HPN was estimated to realise monthly savings of $4,860 per patient (95% CI 2,700–7,000) compared with provision of PN in hospital, with even greater savings in elderly patients and those with underlying malignancy.³⁵⁹
• ESPGHAN–ESPEN guidelines on HPN in children note a paediatric study (age range 0.04–15.83 years at start of hospital TPN) which reported that the number of septic episodes was significantly reduced when children were transferred to HPN (from 1/142 days in hospital [interquartile range 99–290] to 1/567 days at home [251–614); based on 17,562 and 10,348 hospital-days, respectively). The cost of treating an episode of sepsis was estimated at £4,733–6,495. Overall HPN was associated with potential savings of approximately €1 million in a single year.³⁶⁰
• A systematic review and meta-analysis of five studies comparing HPN costs to hospital-based PN found that HPN was significantly more cost-effective. The expenditure for HPN was between 60% and 76% lower compared to hospital-based PN, highlighting the substantial cost savings associated with home care versus hospital care for patients requiring parenteral nutrition.³⁶¹

HPN personal and productivity costs

• An observational, retrospective study aimed to assess the total costs of HPN from a societal perspective, considering not only direct healthcare costs but also personal and productivity costs. Data were collected from 22 adult patients receiving HPN for more than 3 months at Gregorio Marañón University Hospital in Madrid, Spain. The study found that personal costs averaged €729.49 per patient annually, with productivity costs at €256.39. The total cost of HPN was €14,460.87 per patient per year, with direct healthcare and non-healthcare costs comprising 96.46% of the total. Personal and productivity costs contributed 2.62% and 0.92%, respectively. These findings highlight that while direct healthcare costs dominate, personal and productivity costs should also be considered in the economic evaluation of HPN. ³⁶²
• From 2010 to 2020, the reimbursement cost for adult HPN in particular exceeded €146 million, with total health-related reimbursements for HPN patients nearing €242 million. On average, HPN costs accounted for 61% of the total healthcare costs for HPN patients. During the observation period, both HPN and total healthcare reimbursement costs increased significantly, with a 2.6-fold and 2.57-fold rise, respectively. The number of patient days also increased 2.9-fold. Interestingly, the cost of HPN per patient decreased over the years, with an estimated reduction of €209,478 (95% CI: -€303,436 to -€115,521, p-trend < 0.001), as did the total healthcare costs per HPN patient (estimate -€184,008, 95% CI: -€296,679 to -€71,337). In comparison, the average cost of healthcare per citizen in Poland was 36.8 times lower than the healthcare costs for HPN patients, highlighting the significant financial burden that HPN places on healthcare systems. ²⁴⁷

• A 1996 cost–utility analysis by Richards and Irving, taking the perspective of the UK National Health Service, reported that HPN was 65% more cost-effective than hospital care in patients with IF. This was based on detailed data from 64 patients from a single hospital who received HPN for a median of 6 nights per week (range 2–7) for a median of 4 years (range 0.5–15). HPN accounted for 77% of the total cost in the first year (£34,157 of £44,288, which included costs for equipment that would not be incurred in subsequent years). QALYs gained were 0.516 in year 1, 2.58 in year 5, and 5.16 in year 10, with marginal costs per QALY gained of £85,829, £66,224, and £54,734, respectively. Hospitalisation reduces utility, and a patient treated in hospital for the median 4 years would incur costs of £312,595, yielding a cost per QALY of £189,451. HPN is life- saving for many patients. The quality-adjusted survival for younger patients (<44 years) is significantly better than that for older patients (>55 years, which significantly reduces the marginal cost per QALY. Furthermore, the longer a patient survives, the more cost- effective HPN becomes. Weaning from HPN because of intestinal adaptation reduces the cost per QALY even further. The results of this cost–utility analysis support timely initiation of HPN in patients with cancer, that is, when performance status, which is an independent predictor of survival, is higher.
• A meta-analysis by Arship et al. (2021) ³⁶¹ showed that the annual cost of HPN in adult patients across Europe ranged widely—from €8,000 to €77,000, depending on the pricing year and healthcare setting. Importantly, the cost of HPN significantly decreased over time. According to Canovai et al., ³⁶³ HPN costs dropped by 15% in the second year and 22% in the third year compared to the first year. By the fifth year, the annual cost was approximately 40% lower than in the initial year, highlighting a clear downward trend in long-term HPN expenses. This cost reduction was mainly attributed to fewer hospital admissions and HPN-related complications over time. ³⁶¹