Pith. sign in

REVIEW 3 major objections 4 minor 101 references

Many happy returns: machine learning to support platelet issuing and waste reduction in hospital blood banks

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A machine-learning model that predicts which platelet requests will be returned can guide hospital issuing decisions to cut waste without lowering service levels.

desk verdict Novel returns-aware platelet issuing policy, but the advertised 14% saving comes from an idealized simulation; the paper's own real-demand evaluation shows a smaller, still positive effect. read the letter →

arxiv 2411.14939 v1 pith:RY43ZXJ7 submitted 2024-11-22 cs.LG

Joseph Farrington , Samah Alimam , Martin Utley , Kezhi Li , Wai Keong Wong This is my paper · ORCID
classification cs.LG MSC 68T0590B05
keywords plateletinventoryissuingpolicymachinelearningreturnswastagereductionsimulation-firstevaluationhospitalbloodbankgradient-boostedtrees
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that the routine practice of issuing the oldest platelet unit first becomes suboptimal when some issued units are returned unused, and that a machine-learning model can exploit returns to reduce waste. It trains a model on 17,297 platelet requests to predict, for each request, whether at least one issued unit will come back, reaching an AUROC of 0.74 on 9,353 held-out requests. Embedding that model in a simulation of a hospital blood bank, the paper estimates that a policy of issuing the freshest units when returns are predicted and the oldest otherwise would cut wastage by about 14% (from 0.91% to 0.78%) at the partner hospital without lowering the service level. The estimated benefit grows when return rates are higher or remaining shelf life on arrival is shorter, as in a US-style scenario where wastage falls from 5.0% to 4.3% in the real-demand evaluation.

What carries the argument

The load-bearing mechanism is the YUPR issuing policy combined with a pre-deployment simulation that maps combinations of model sensitivity and specificity to wastage and service level. The simulation models daily ordering, morning and afternoon demand, returns at midday, slippage (units returned in a state that cannot be reissued), and expiry, with parameters estimated from hospital data. A single-unit demand assumption lets each request's true return label be drawn as a Bernoulli($\rho$) and the predicted label from fixed sensitivity $\alpha$ and specificity $\beta$, producing contour plots that define the region where YUPR beats OUFO; the trained model's ROC curve is then superimposed to read off the wastage reduction. The ML model itself is a gradient-boosted decision tree trained with features such as recent platelet count and how far in advance the request was made.

What would settle it

Run YUPR against Oldest-Unit-First-Out prospectively at a hospital with a return rate around 8% and measure wastage and service level over a full year; if wastage is not lower, or the service level falls, the central claim fails. The paper's own real-demand simulation already provides a partial check, showing smaller gains (1.1% versus 1.2% waste at the partner hospital) than the 14% headline figure.

Watch

Extended reading notes

Core claim

The central claim is that platelet returns are predictable enough to change issuing decisions, and that acting on the prediction improves the blood bank's key performance indicators. The paper proposes Youngest Unit for Predicted Returns (YUPR): if the model predicts a request will not be transfused, issue the freshest unit so that a returned unit has maximum time left before expiry; otherwise issue the oldest unit as usual. In the simulation, this policy beats Oldest-Unit-First-Out on wastage with no service-level loss, and the advantage is largest with higher return rates and shorter remaining shelf life. The paper also claims a methodological point: a simulation-first evaluation can show whether a predictive model has operational value before the model is built, and can guide the performance requirements for the model.

Load-bearing premise

The headline 14% reduction comes from contours built on the assumption that a model is fully captured by fixed sensitivity and specificity and that returns happen independently with a fixed probability; if returns and model errors correlate with wards, timing, platelet counts, or stock state, the contour estimate can misstate the true benefit.

Editorial extensions

If this is right

  • Hospitals can lower platelet wastage without reducing service by switching issuing from oldest-first to model-guided.
  • The policy matters most where return rates are high and units arrive with less remaining shelf life, because those settings show larger absolute reductions.
  • Simulation-first evaluation can be used to decide whether to invest in ML for inventory workflows before building models, and to set target performance levels.
  • A rule-based or simpler predictor based on the model's insights might capture part of the benefit, since the partner hospital is exploring that route.
  • The benefits require the ability to reissue returned units; sites where returned units are not in reissuable state see no gain.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's mechanism is not limited to platelets: the same 'issue freshest when return is likely' logic could apply to red blood cells or other perishable medical stock where units are returned and reissued, though the paper only studies platelets.
  • The paper's simulation treats model quality by fixed sensitivity and specificity with independent return labels, so a richer evaluation that lets return probability depend on ward, timing, platelet count, and stock state would likely change the estimated benefit; the paper's own real-demand simulation shows a smaller benefit at the partner hospital (1.1% versus 1.2% waste).
  • Combining return predictions with replenishment decisions, such as ordering less when predicted returns are high, could reduce waste further; the paper notes this is not tested.
  • The wastage-minimising classification threshold depends on the operating scenario, so hospitals adopting the policy would need to recalibrate the threshold locally.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes an ML-guided platelet issuing policy (YUPR: youngest unit first for predicted returns) and evaluates it at a single UK hospital trust. An XGBoost model is trained on 17,297 UCLH platelet requests from 2015-2016 and tested on 9,353 requests from 2017, achieving AUROC 0.74. The authors use a simulation-first approach: they build a discrete-event simulation of the blood bank workflow with returns, generate wastage/service-level contour plots over sensitivity/specificity pairs, and then use the test-set sensitivity/specificity at a threshold chosen on the training set to claim a 14% wastage reduction (0.91% to 0.78%) with no service-level detriment. They also provide a complementary real-demand evaluation in Supplementary Note J that plugs actual 2017 requests and model predictions into the same workflow, reporting UCLH wastage 1.2% to 1.1% and R&R shelf-life wastage 5.0% to 4.3%, again with no service-level reduction. Sensitivity analyses show that the potential benefit increases with higher return rates and shorter remaining shelf life on arrival.

Significance. If the qualitative claim holds, this is a useful contribution to blood inventory management: it identifies post-issue returns as a largely neglected factor, proposes a concrete issuing policy that targets returns, and demonstrates that even an imperfect predictive model can reduce wastage while maintaining service level. The core ML evaluation is not circular: the model is trained on independent historical labels and evaluated on a temporal holdout. Additional strengths include the simulation-first methodology, the use of KPI-based evaluation rather than only AUROC, the inclusion of real-demand evaluation in the supplement, and the provided code and non-identifiable simulation inputs. The main weakness is that the headline quantitative claim is taken from an idealized contour-lookup procedure rather than from the paper's own more realistic real-demand evaluation, and that realistic evaluation is reported without uncertainty. The direction of the effect is consistent across both evaluation routes, but the magnitude in the abstract is not robust to the paper's more realistic route.

major comments (3)
  1. [Abstract; Section 2.2; Supplementary Note J] The abstract and Section 2.2 headline a 14% reduction in wastage (0.91% to 0.78%) for the UCLH scenario. This figure is obtained by reading the test-set sensitivity/specificity off contour plots produced under the simplifying assumptions of Section 4.3.3 (single-unit requests, independent Bernoulli return labels, fixed sensitivity/specificity). The paper's own real-demand evaluation in Supplementary Note J gives UCLH wastage 1.2% under OUFO and 1.1% under the ML-guided policy, i.e., an 8.3% relative reduction rather than 14%, and 5.0% to 4.3% under R&R shelf life; no confidence interval is reported for either one-year point estimate. This discrepancy is material because the abstract presents 14% as the headline estimate. The authors should either make the Note J result the primary estimate with an uncertainty quantification, or clearly label the 14% figure as an idealized single-unit simulation result, not the estimated real-world benefit.
  2. [Section 4.2.2] The classification threshold is selected by minimizing estimated wastage on the training set using the same contour plots from Experiments 2 and 4, and the test-set sensitivity/specificity at that threshold are then looked up on those same contours. Because the contours are generated with inputs from UCLH 2015-2016 (or R&R), the reported KPI is an in-sample estimate of an optimized operating point and has no uncertainty estimate. A reader cannot tell whether the 1.1% versus 1.2% difference observed in Supplementary Note J is within sampling noise. Please provide a confidence interval or bootstrap for the real-demand KPI simulation and, if possible, a pre-registered or less simulation-dependent threshold selection rule.
  3. [Section 4.3.3; Algorithm 1] Algorithm 1 assumes that the true return label is an independent Bernoulli(rho) draw and that the model's prediction depends only on the true label through fixed sensitivity and specificity. This removes any correlation between returns, model errors, and request features such as ward, platelet count, or stock state, correlations that the SHAP analysis itself shows to be important. The manuscript acknowledges the single-unit discrepancy in Section 4.2.2 and supplies Supplementary Note J as a corrective, but the corrected route is a single one-year point estimate. To make the headline claim load-bearing, the authors should quantify the gap between the two routes, for example by running the real-demand workflow over multiple years or by reporting both the idealized and real-demand results with appropriate caveats.
minor comments (4)
  1. [Figure 1 caption] The caption says 'combinations of sensitivity and sensitivity' and should read 'sensitivity and specificity'.
  2. [Section 4.3.4 and Figure E2 caption] There are duplicated words: 'using a a standing order' in Section 4.3.4 and 'on the the daily cost' in the Figure E2 caption.
  3. [Discussion and Section 4.2.2] The Discussion says 'impact on a patent's health' and should say 'patient's health'; Section 4.2.2 contains 'a region of the of the curve'.
  4. [Section 2.2] The statement 'no detriment to service level' is supported by the real-demand Note J numbers (99.4% vs 99.4% and 99.1% vs 99.1%), but the contour-interpolation route used for the headline 14% figure does not report a service-level value at the selected operating point; clarifying which route supports each claim would improve transparency.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the ML model is evaluated on a temporal holdout and the KPI claims are cross-checked by an independent real-demand simulation; the headline 14% versus Supplementary Note J discrepancy is a model-assumption gap, not a circular reduction.

full rationale

The paper's central ML result is not circular: the model is trained on 17,297 requests from 2015-2016 and evaluated on 9,353 held-out requests from 2017, with AUROC 0.74 reported on that holdout (Sections 2.1 and 4.2.1). The operational benefit is estimated through two routes. The first route reads the test-set sensitivity and specificity off wastage contours generated by Experiments 2 and 4; the second, described in Supplementary Note J, plugs real 2017 demand and model predictions directly into the simulated workflow and reports wastage of 1.1% versus 1.2% at UCLH and 4.3% versus 5.0% under the US shelf-life distribution. Neither route reduces to a fitted quantity by construction: the simulation contours are built from observed rates (rho, phi, shelf-life distributions) and an explicit grid of hypothetical-model sensitivities and specificities, and the trained model's operating point is measured on held-out requests rather than tuned to the reported KPI. The classification threshold is selected on the training set to minimize wastage estimated from the same contours, but the final test metrics are independent, so this is standard threshold validation rather than circularity. The paper explicitly acknowledges that the contour route assumes single-unit requests and independent return labels (Section 4.2), and the real-demand simulation is an honest, more realistic check; the difference between the 14% headline and the smaller UCLH real-demand benefit is a robustness and reporting concern, not a definitional identity. There is a minor self-citation to the authors' prior work [93] to justify using an (s,S) replenishment policy and simulation-optimization fitting, but this is methodological and not load-bearing: the policies are described as simple baselines and their parameters are refit in the present study. Overall, no circular step meeting the required evidentiary standard was found.

Assumptions & free parameters 7 free parameters · 8 assumptions · 0 invented entities

The simulation that generates the headline wastage reduction is parameterized by UCLH 2015-2016 return rate, slippage, demand, and shelf-life distributions; replenishment parameters are fit by simulation optimization; the ML threshold is chosen to minimize simulated wastage. Several workflow simplifications (Poisson demand, independent Bernoulli returns, no blood-type matching, zero lead time) are necessary for tractability and are acknowledged in the text. No new physical entities are introduced.

free parameters (7)
  • Return rate rho = 0.08
    Estimated from UCLH 2015-2016 requests as the proportion of issued units not transfused; central to the simulation's return flow.
  • Slippage rate phi = 0.07
    Upper-limit estimate from UCLH data; sensitivity analysis shows the benefit of YUPR vanishes at high slippage.
  • Mean daily demand by weekday, mu_tau = Table C1: 28.8, 33.4, 26.2, 28.4, 30.8, 18.6, 19.6
    Estimated from UCLH 2015-2016 total platelet requests including returns; used to sample Poisson demand in simulation.
  • Remaining useful life on arrival distribution, Delta_tau = Table C3 (UCLH) and Table C5 (sensitivity); R&R 3-day distribution
    Multinomial probabilities per weekday estimated from UCLH data, plus a US-hospital scenario from Rajendran and Ravindran [73].
  • Replenishment policy parameters (s_tau, S_tau) = Fit per scenario by Optuna simulation optimization
    Fourteen parameters per weekday, refit for each sensitivity/specificity pair in Experiments 1-4 and for each setting in sensitivity analyses.
  • Classification threshold = Training-set threshold minimizing simulated wastage
    Selected on the training set using the simulation's wastage contours; the test-set operating point used for the reported benefit depends on this choice.
  • XGBoost hyperparameters = Final values in Table I15, e.g., max_depth=6, learning_rate=0.11, scale_pos_weight=3.15
    Tuned by Optuna over 200 trials using 10-fold CV and partial-AUROC; these values affect the trained model's predictions.
assumptions (8)
  • domain assumption Demand follows a Poisson distribution, split equally between morning and afternoon.
    Used in stages 2 and 4 of the simulated workflow (Section 4.3.1, equations B6-B12); not verified against the empirical demand distribution.
  • domain assumption In the initial simulation, return labels are independent Bernoulli draws with probability rho, regardless of request features.
    Algorithm 1 in Supplementary Note B samples true labels from Bernoulli(rho); this makes the model's value depend only on sensitivity/specificity and ignores feature-conditional return patterns.
  • domain assumption Returned units can be reissued if they have not expired and are not lost to slippage.
    The YUPR policy's benefit depends on reissuability; the authors state this is valid at UCLH due to remote agitators but may not hold elsewhere (Discussion).
  • domain assumption All platelet units are interchangeable; no ABO compatibility or special patient requirements are modeled.
    Explicitly acknowledged as a limitation in the Discussion; real issuing decisions must respect blood group and special requirements.
  • domain assumption There is no clinical reason to prefer fresher platelets for specific patients.
    Assumed in Section 4.3.3 and discussed with reference to a systematic review [79] finding no storage-time outcome relationship.
  • domain assumption Routine orders have zero lead time and arrive immediately.
    Stated in Supplementary Note B.1; real deliveries arrive in scheduled slots, which the simulation abstracts away.
  • domain assumption Cost parameters from Rajendran and Ravindran [73] apply to the UCLH setting.
    Holding, shortage, wastage, and ordering costs are taken from a US regional medical centre rather than estimated at UCLH.
  • domain assumption The test-set sensitivity and specificity at the selected threshold transfer to the simulated workflow's operating point.
    Section 4.2.2 looks up wastage from simulation contours using test-set metrics; this assumes the model's error rates are constant across workflow states.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Many happy returns: machine learning to support platelet issuing and waste reduction in hospital blood banks." pith.science (2026). https://pith.science/paper/RY43ZXJ7

@misc{pith2026241114939,
  author       = {Pith},
  title        = {Pith review of: Many happy returns: machine learning to support platelet issuing and waste reduction in hospital blood banks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RY43ZXJ7}},
  note         = {Machine review of arXiv:2411.14939}
}
read the original abstract

Efforts to reduce platelet wastage in hospital blood banks have focused on ordering policies, but the predominant practice of issuing the oldest unit first may not be optimal when some units are returned unused. We propose a novel, machine learning (ML)-guided issuing policy to increase the likelihood of returned units being reissued before expiration. Our ML model trained to predict returns on 17,297 requests for platelets gave AUROC 0.74 on 9,353 held-out requests. Prior to ML model development we built a simulation of the blood bank operation that incorporated returns to understand the scale of benefits of such a model. Using our trained model in the simulation gave an estimated reduction in wastage of 14%. Our partner hospital is considering adopting our approach, which would be particularly beneficial for hospitals with higher return rates and where units have a shorter remaining useful life on arrival.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

101 extracted references · 80 canonical work pages

  1. [1]

    Flint, A. W. et al. Is platelet expiring out of date? A systematic review.Transfus. Med. Rev. 34, 42–50 (2020)

    2020

  2. [2]

    NHSBT hospital blood supply chain annual report 2017/18

    Blood Stocks Management Scheme. NHSBT hospital blood supply chain annual report 2017/18. Tech. Rep. (2018). https://nhsbtdbe.blob.core.windows.net/ umbraco-assets-corp/15951/nhsbt-annual-report-2017-18.pdf

    2018

  3. [3]

    2023 BSMS 10 year component review

    Blood Stocks Management Scheme. 2023 BSMS 10 year component review. Tech. Rep. (2023). https://nhsbtdbe.blob.core.windows.net/umbraco-assets-corp/ 31272/bsms-10-year-component-review-2.pdf

    2023

  4. [4]

    & Cotton, S

    Yates, N., Stanger, S., Wilding, R. & Cotton, S. Approaches to assessing and minimizing blood wastage in the hospital and blood supply chain. ISBT Sci. Ser. 12, 91–98 (2017)

    2017

  5. [5]

    H., Yates, N., Wilding, R

    Stanger, S. H., Yates, N., Wilding, R. & Cotton, S. Blood inventory management: hospital best practice. Transfus. Med. Rev. 26, 153–163 (2012)

    2012

  6. [6]

    & Garraud, O

    Tissot, J.-D. & Garraud, O. Ethics and blood donation: A marriage of convenience. Presse Med. 45, e247–e252 (2016)

    2016

  7. [7]

    Data driven personalized management of hospital inventory of perishable and substitutable blood products

    Dumkreiger, G. Data driven personalized management of hospital inventory of perishable and substitutable blood products . Ph.D. thesis, Arizona State University, Phoenix, AZ, USA (2020). https://keep.lib.asu.edu/items/158661

    2020

  8. [8]

    F., Brailsford, S

    Osorio, A. F., Brailsford, S. C. & Smith, H. K. A structured review of quantitative models in the blood supply chain: a taxonomic framework for decision-making. Int. J. Prod. Econ. 53, 7191–7212 (2015)

    2015

Show all 101 references
  1. [9]

    Pirab´ an, A., Guerrero, W. J. & Labadie, N. Survey on blood supply chain management: models and methods. Comput. Oper. Res. 112 (2019)

    2019

  2. [10]

    & Barbosa-P´ ovoa, A

    Torrado, A. & Barbosa-P´ ovoa, A. Towards an Optimized and Sustainable Blood Supply Chain Network under Uncertainty: A Literature Review. Clean. Logist. Supply Chain 3, 100028 (2022)

    2022

  3. [11]

    Optimal ordering, issuance and disposal policies for inventory management of perishable products

    Haijema, R. Optimal ordering, issuance and disposal policies for inventory management of perishable products. Int. J. Prod. Econ. 157, 158–169 (2014)

    2014

  4. [12]

    H., Wilding, R., Yates, N

    Stanger, S. H., Wilding, R., Yates, N. & Cotton, S. What drives perishable inventory management performance? Lessons learnt from the UK blood supply chain. Supply Chain Manag. 17, 107–123 (2012)

    2012

  5. [13]

    & Teunter, R

    Bakker, M., Riezebos, J. & Teunter, R. H. Review of inventory systems with deterioration since 2001. Eur. J. Oper. Res. 221, 275–284 (2012). 20

    2012

  6. [14]

    & Routroy, S

    Chaudhary, V., Kulshrestha, R. & Routroy, S. State-of-the-art literature review on inventory models for perishable products. J. Adv. Manag. Res. 15, 306–346 (2018)

    2018

  7. [15]

    & Sauer, J

    Janssen, L., Claus, T. & Sauer, J. Literature review of deteriorating inventory models by key topics from 2012 to 2015. Int. J. Prod. Econ. 182, 86–112 (2016)

    2016

  8. [16]

    Z., Scheller–Wolf, A

    Karaesmen, I. Z., Scheller–Wolf, A. & Deniz, B. in Managing Perishable and Aging Inventories: Review and Future Research Directions (eds Kempf, K. G., Keskinocak, P. & Uzsoy, R.) Planning Production and Inventories in the Extended Enterprise: A State of the Art Handbook, Volum...

    2011

  9. [17]

    & Solina, V

    Mirabelli, G. & Solina, V. Optimization strategies for the integrated manage- ment of perishable supply chains: A literature review. J. Ind. Eng. Manag. 15, 58–91 (2022)

    2022

  10. [18]

    Perishable inventory theory: a review

    Nahmias, S. Perishable inventory theory: a review. Oper. Res. 30, 680–708 (1982)

    1982

  11. [19]

    Pierskalla, W. P. & Roach, C. D. Optimal Issuing Policies for Perishable Inventory. Manage. Sci. 18, 603–614 (1972)

    1972

  12. [20]

    & Alkhaledi, K

    Dalalah, D., Bataineh, O. & Alkhaledi, K. A. Platelets inventory management: a rolling horizon sim–opt approach for an age-differentiated demand. J. Simul. 13, 209–225 (2019)

    2019

  13. [21]

    & Scheller-Wolf, A

    Civelek, I., Karaesmen, I. & Scheller-Wolf, A. Blood platelet inventory management with protection levels. Eur. J. Oper. Res. 243, 826–838 (2015)

    2015

  14. [22]

    & Abbasi, B

    Dillon, M., Oliveira, F. & Abbasi, B. A two-stage stochastic programming model for inventory management in the blood supply chain. Int. J. Prod. Econ. 187, 27–41 (2017)

    2017

  15. [23]

    & Liao, T

    Duan, Q. & Liao, T. W. Optimization of blood supply chain with shortened shelf lives and ABO compatibility. Int. J. Prod. Econ. 153, 113–129 (2014)

    2014

  16. [24]

    & Barbosa-P´ ovoa, A

    Meneses, M., Marques, I. & Barbosa-P´ ovoa, A. Blood inventory management: ordering policies for hospital blood banks under uncertainty. Int. T. Oper. Res. (2021)

    2021

  17. [25]

    & Hosseinifard, S

    Abbasi, B. & Hosseinifard, S. Z. On the Issuing Policies for Perishable Items such as Red Blood Cells and Platelets in Blood Service. Decision Sci. 45, 995–1020 (2014)

    2014

  18. [26]

    P., Fontaine, M

    Atkinson, M. P., Fontaine, M. J., Goodnough, L. T. & Wein, L. M. A novel allocation strategy for blood transfusions: investigating the tradeoff between the 21 age and availability of transfused blood. Transfusion 52, 108–117 (2012)

    2012

  19. [27]

    & Wahab, M

    Abdulwahab, U. & Wahab, M. I. Approximate dynamic programming modeling for a typical blood platelet bank. Comput. Ind. Eng. 78, 259–270 (2014)

    2014

  20. [28]

    total solution

    Staves, J. et al. Electronic remote blood issue: a combination of remote blood issue with a system for end-to-end electronic control of transfusion to pro- vide a “total solution” for a safe and timely hospital blood transfusion service. Transfusion 48, 415–424 (2008)

    2008

  21. [29]

    Jennings, J. B. An Analysis of Hospital Blood Bank Whole Blood Inventory Control Policies. Transfusion 8, 335–342 (1968)

    1968

  22. [30]

    Blood inventory management in the type and screen era

    Pereira, A. Blood inventory management in the type and screen era. Vox Sang. 89, 245–250 (2005)

    2005

  23. [31]

    Rytil¨ a, J. S. & Spens, K. M. Using simulation to increase efficiency in blood supply chains. Manag. Res. News 29, 801–819 (2006)

    2006

  24. [32]

    & Brailsford, S

    Katsaliaki, K. & Brailsford, S. C. Using simulation to improve the blood supply chain. J. Oper. Res. Soc. 58, 219–227 (2007)

    2007

  25. [33]

    P., Rimm, A

    Yahnke, D. P., Rimm, A. A., Mundt, C. J., Aster, R. H. & Hurst, T. M. Analysis and Optimization of a Regional Blood Bank Distribution Process. Transfusion 12, 111–118 (1972)

    1972

  26. [34]

    P., Rimm, A

    Yahnke, D. P., Rimm, A. A., Makowski, G. G. & Aster, R. H. Analysis and Optimization of a Regional Blood Bank Distribution Process: II. Derivation and Use of a Method for Evaluating Hospital Management Procedures. Transfusion 13, 156–169 (1973)

    1973

  27. [35]

    Britten, A. F. H. & Geurtze, D. G. Weekly Rotation of Blood Inventory—A System for Supplying Small Hospitals. Transfusion 19, 738–741 (1979)

    1979

  28. [36]

    & Xiao, T

    Chen, K., Song, J.-S., Shang, J. & Xiao, T. Managing hospital platelet inventory with mid-cycle expedited replenishments and returns. Prod. Oper. Manag. 31, 2015–2037 (2022)

    2022

  29. [37]

    & Sadeghi, S

    Ahmadimanesh, M., Pooya, A., Safabakhsh, H. & Sadeghi, S. Designing an Optimal Model of Blood Logistics Management with the Possibility of Return in the Three-Level Blood Transfusion Network. Research Square (2022). https: //www.researchsquare.com/article/rs-1187827/v1

    2022

  30. [38]

    & Bilolikar, V

    Ambilkar, P., Dohale, V., Gunasekaran, A. & Bilolikar, V. Product returns management: a comprehensive review and future research agenda. Int. J. Prod. Econ. 0, 1–25 (2021). 22

    2021

  31. [39]

    & Abbey, J

    Abdulla, H., Ketzenberg, M. & Abbey, J. D. Taking stock of consumer returns: A review and classification of the literature. J. Oper. Manag. 65, 560–605 (2019)

    2019

  32. [40]

    & Wang, F.-Y

    Canda, A., Yuan, X.-M. & Wang, F.-Y. Modeling and forecasting product returns: An industry case study. 2015 IEEE International Conference on Indus- trial Engineering and Engineering Management (IEEM) 871–875 (Singapore, 6-9 December 2015)

    2015

  33. [41]

    Tsiliyannis, C. A. Markov chain modeling and forecasting of product returns in remanufacturing based on stock mean-age. Eur. J. Oper. Res. 271, 474–489 (2018)

    2018

  34. [42]

    C., Sim, C.-K

    Chou, M. C., Sim, C.-K. & Yuan, X.-M. Policies for inventory models with product returns forecast from past demands and past sales. Ann. Oper. Res. 288, 137–180 (2020)

    2020

  35. [43]

    & Ward, A

    Cui, H., Rajagopalan, S. & Ward, A. R. Predicting product return volume using machine learning methods. Eur. J. Oper. Res. 281, 612–627 (2020)

    2020

  36. [44]

    & Morley, J

    Joshi, I. & Morley, J. Artificial Intelligence: How to get it right. NHS Transforma- tion Directorate (2019). https://transform.england.nhs.uk/media/documents/ NHSX AI report.pdf

    2019

  37. [45]

    Zhu, Y., Li, J., He, J., Quanz, B. L. & Deshpande, A. A. A Local Algorithm for Product Return Prediction in E-Commerce. Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence 3718–3724 (Stockholm, Sweden, July 13-19 2018)

    2018

  38. [46]

    Li, X., Zhuang, Y., Fu, Y. & He, X. A trust-aware random walk model for return propensity estimation and consumer anomaly scoring in online shopping. Sci. China Inform. Sci. 62, 52101 (2019)

    2019

  39. [47]

    Fu, Y. et al. Fused latent models for assessing product return propensity in online commerce. Decis. Support Syst. 91, 77–88 (2016)

    2016

  40. [48]

    & Prastacos, G

    Brodheim, E. & Prastacos, G. Demand, usage and issuing of blood at hospital blood banks. Tech. Rep., Operations Research Laboratory, The New York Blood Center (1980)

    1980

  41. [49]

    H., Milstein, A

    Shah, N. H., Milstein, A. & Bagley, P., Steven C. Making Machine Learning Models Clinically Useful. JAMA, J. Am. Med. Assoc. 322, 1351–1352 (2019)

    2019

  42. [50]

    & Peng, S

    Zhou, Q., Chen, Z.-h., Cao, Y.-h. & Peng, S. Clinical impact and quality of ran- domized controlled trials involving interventions evaluating artificial intelligence prediction tools: a systematic review. npj Digit. Med. 4, 1–12 (2021). 23

    2021

  43. [51]

    C., Asch, S

    Li, R. C., Asch, S. M. & Shah, N. H. Developing a delivery science for artificial intelligence in healthcare. npj Digit. Med. 3, 1–3 (2020)

    2020

  44. [52]

    V., Rajaram, K

    Miˇ si´ c, V. V., Rajaram, K. & Gabel, E. A simulation-based evaluation of machine learning models for clinical decision support: application and analysis using hospital readmission. npj Digit. Med. 4, 1–11 (2021)

    2021

  45. [53]

    & Shah, N

    Wornow, M., Gyang Ross, E., Callahan, A. & Shah, N. H. APLUS: A Python library for usefulness simulations of machine learning models in healthcare. J. Biomed. Inform. 139, 104319 (2023)

    2023

  46. [54]

    Taylor, J. A. et al. The road to hell is paved with good intentions: the experience of applying for national data for linkage and suggestions for improvement. BMJ Open 11, e047575 (2021)

    2021

  47. [55]

    Q., Behdani, B

    Viet, N. Q., Behdani, B. & Bloemhof, J. The value of information in supply chain decisions: A review of the literature and research agenda. Comput. Ind. Eng. 120, 68–82 (2018)

    2018

  48. [56]

    & Carter, M

    Kopach, R., Balcioˇ glu, B. & Carter, M. Tutorial on constructing a red blood cell inventory management system with two demand rates. Eur. J. Oper. Res. 185, 1051–1059 (2008)

    2008

  49. [57]

    & Nagarur, N

    Yuzgec, E., Han, Y. & Nagarur, N. A simulation model for blood supply chain systems. Proceedings of the 2013 Industrial and Systems Engineering Research Conference 1703–1711 (San Juan, Puerto Rico, 18-22 May 2013)

    2013

  50. [58]

    J., Carter, C

    Katz, A. J., Carter, C. W., Saxton, P., Blutt, J. & Kakaiya, R. M. Simulation analysis of platelet production and inventory management. Vox Sang. 44, 31–36 (1983)

    1983

  51. [59]

    & Ettkin, L

    Asllani, A., Culler, E. & Ettkin, L. A simulation-based apheresis platelet inventory management model. Transfusion 54, 2730–2735 (2014)

    2014

  52. [60]

    & Bast ´ ıas, A

    Baesler, F., Nemeth, M., Mart ´ ınez, C. & Bast ´ ıas, A. Analysis of inventory strate- gies for blood components in a regional blood center using process simulation. Transfusion 54, 323–330 (2014)

    2014

  53. [61]

    & Srinivas, S

    Rajendran, S. & Srinivas, S. Hybrid ordering policies for platelet inventory management under demand uncertainty. IISE Trans. Healthc. Syst. Eng. 10, 113–126 (2020)

    2020

  54. [62]

    & Liao, T

    Duan, Q. & Liao, T. W. A new age-based replenishment policy for supply chain inventory optimization of highly perishable products. Int. J. Prod. Econ. 145, 658–671 (2013). 24

    2013

  55. [63]

    Blake, J. T. et al. Optimizing the platelet supply chain in Nova Scotia. Pro- ceedings of the 29th meeting of the European Working Group on Operational Research Applied to Health Services 47–66 (Prague, Czech Republic, 27 July - 1 August 2003)

    2003

  56. [64]

    & van Dijk, N

    Haijema, R., van der Wal, J. & van Dijk, N. M. Blood platelet production: optimization by dynamic programming and simulation. Comput. Oper. Res. 34, 760–779 (2007)

    2007

  57. [65]

    Li, N., Chiang, F., Down, D. G. & Heddle, N. M. A decision integration strategy for short-term demand forecasting and ordering for red blood cell components. Oper. Res. Health Care 29, 100290 (2021)

    2021

  58. [66]

    Motamedi, M., Li, N., Down, D. G. & Heddle, N. M. Demand forecasting for platelet usage: from univariate time series to multivariate models. arXiv (2021). http://arxiv.org/abs/2101.02305

    2021 arXiv

  59. [67]

    & Spreckelsen, C

    Schilling, M., Rickmann, L., Hutschenreuter, G. & Spreckelsen, C. Reduction of platelet outdating and shortage by forecasting demand with statistical learning and deep neural networks: modeling study. JMIR Med. Inform. 10, e29978 (2022)

    2022

  60. [68]

    Guan, L. et al. Big data modeling to predict platelet usage and minimize wastage in a tertiary care system. P. Natl. Acad. Sci. USA 114, 11368–11373 (2017)

    2017

  61. [69]

    & Sarhangian, V

    Abouee-Mehrizi, H., Mirjalili, M. & Sarhangian, V. Data-driven platelet inven- tory management under uncertainty in the remaining shelf life of units. Prod. Oper. Manag. 31, 3914–3932 (2022)

    2022

  62. [70]

    & Kingsman, B

    Fildes, R. & Kingsman, B. Incorporating demand uncertainty and forecast error in supply chain planning models. J. Oper. Res. Soc. 62, 483–500 (2011)

    2011

  63. [71]

    & Jodlbauer, H

    Altendorfer, K., Felberbauer, T. & Jodlbauer, H. Effects of forecast errors on optimal utilisation in aggregate production planning with stochastic customer demand. Int. J. Prod. Econ. 54, 3718–3735 (2016)

    2016

  64. [72]

    & Graman, G

    Sanders, N. & Graman, G. Quantifying costs of forecast errors: A case study of the warehouse environment. Omega 37, 116–125 (2009)

    2009

  65. [73]

    & Ravindran, A

    Rajendran, S. & Ravindran, A. R. Platelet ordering policies at hospitals using stochastic integer programming model and heuristic approaches to reduce wastage. Comput. Ind. Eng. 110, 151–164 (2017)

    2017

  66. [74]

    & Lee, S.-I

    Lundberg, S. & Lee, S.-I. A unified approach to interpreting model predic- tions. Proceedings of the 31st International Conference on Neural Information Processing Systems 4765–4774 (Long Beach, CA, USA, 4–9 December 2017). 25

    2017

  67. [75]

    Kron, A. et al. Multicenter observational study evaluating the impact of platelet transport bags on product wastage. Transfusion 61, 1383–1388 (2021)

    2021

  68. [76]

    Fenwick, E. et al. Value of Information Analysis for Research Decisions—An Introduction: Report 1 of the ISPOR Value of Information Analysis Emerging Good Practices Task Force. Value Health 23, 139–150 (2020)

    2020

  69. [77]

    & Hudgson, N

    Feizi, A., Orfanoudaki, A., Saghafian, S. & Hudgson, N. Vertical Patient Streaming in Emergency Departments. SSRN (2023). https://papers.ssrn.com/ abstract=4465161

    2023

  70. [78]

    G., Shah, N

    Seneviratne, M. G., Shah, N. H. & Chu, L. Bridging the implementation gap of machine learning in healthcare. BMJ Innov. 6 (2020)

    2020

  71. [79]

    Aubron, C., Flint, A. W. J., Ozier, Y. & McQuilten, Z. Platelet storage duration and its clinical and transfusion outcomes: a systematic review. Crit. Care 22, 185 (2018)

    2018

  72. [80]

    Estcourt, L. J. et al. Guidelines for the use of platelet transfusions. Br. J. Haematol. 176, 365–394 (2017)

    2017

  73. [81]

    Blood transfusion (NICE guideline 24) (2015)

    National Institute for Health and Care Excellence. Blood transfusion (NICE guideline 24) (2015). https://www.nice.org.uk/guidance/ng24/chapter/ Recommendations

    2015

  74. [82]

    & Barty, R

    Blake, J., Heddle, N., Hardy, M. & Barty, R. Simplified platelet ordering using shortage and outdate targets. Int. J. Healthc. Manag. 1, 145–166 (2010)

    2010

  75. [83]

    & Chaudhry, Z

    Sekhar, M., Clark, S., Atugonza, R., Li, A. & Chaudhry, Z. Effective imple- mentation of a patient blood management programme for platelets. Transfusion Med. 26, 422–431 (2016)

    2016

  76. [84]

    & Rezaiee, M

    Abbaspour, A., Jahan, A. & Rezaiee, M. A simple empirical model for blood platelet production and inventory management under uncertainty.J. Amb. Intel. Hum. Comp. 12, 1783–1799 (2021)

    2021

  77. [85]

    Mirjalili, M., Abouee-Mehrizi, H., Barty, R., Heddle, N. M. & Sarhangian, V. A data-driven approach to determine daily platelet order quantities at hospitals. Transfusion 62, 2048–2056 (2022)

    2022

  78. [86]

    & Guestrin, C

    Chen, T. & Guestrin, C. XGBoost: a scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 785–794 (San Francisco, CA, USA, 13–17 August 2016)

    2016

  79. [87]

    & Armon, A

    Shwartz-Ziv, R. & Armon, A. Tabular data: Deep learning is not all you need. Inform. Fusion 81, 84–90 (2022). 26

    2022

  80. [88]

    & Koyama, M

    Akiba, T., Sano, S., Yanase, T., Ohta, T. & Koyama, M. Optuna: A Next- generation Hyperparameter Optimization Framework. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 2623–2631 (Anchorage, AK, USA, August 4-8 2019)

    2019

  81. [89]

    I., Rockette, H

    Ma, H., Bandos, A. I., Rockette, H. E. & Gur, D. On use of partial area under the ROC curve for evaluation of diagnostic performance. Stat. Med. 32, 3449–3458 (2013)

    2013

  82. [90]

    Lange, R. T. gymnax: A JAX-based reinforcement learning environment library, 0.0.5. GitHub (2022). http://github.com/RobertTLange/gymnax

    2022

  83. [91]

    V., Haijema, R., van der Wal, J

    van Dijk, N. V., Haijema, R., van der Wal, J. & Sibinga, C. S. Blood platelet production: a novel approach for practical optimization. Transfusion 49, 411– 420 (2009)

    2009

  84. [92]

    Data-driven modelling and control of hospital blood inventory

    Mirjalili, M. Data-driven modelling and control of hospital blood inventory. Ph.D. thesis, University of Toronto, Toronto, Canada (2022). https://tspace.library. utoronto.ca/bitstream/1807/124976/1/Mirjalili Mahdi 202211 PhD thesis.pdf

    2022

  85. [93]

    Farrington, J., Li, K., Wong, W. K. & Utley, M. Going faster to see further: GPU-accelerated value iteration and simulation for perishable inventory control using JAX. arXiv (2023). http://arxiv.org/abs/2303.10672

    2023 arXiv

  86. [94]

    Blake, J. T. On the use of operational research for managing platelet inventory and ordering. Transfusion 49, 396–401 (2009)

    2009

  87. [95]

    Prioli, K. M. et al. Economic implications of FDA platelet bacterial guidance compliance options: Comparison of single-step strategies. Transfusion 62, 365– 373 (2022)

    2022

  88. [96]

    Szczepiorkowski, Z. M. & Pagano, M. B. Platelet components and bacterial contamination: hospital perspective 2022. Hematology 2022, 430–436 (2022)

    2022

  89. [97]

    van der Meer, P. F. & Korte, D. d. Platelet preservation: Agitation and containers. Transfus. Apher. Sci. 44, 297–304 (2011)

    2011

  90. [98]

    Sutton, R. S. & Barto, A. G. Reinforcement Learning: An Introduction 2 edn (The MIT Press, 2018)

    2018

  91. [99]

    & Zhao, G

    Sun, R. & Zhao, G. Analyses about efficiency of reinforcement learning to supply chain ordering management. IEEE 10th International Conference on Industrial Informatics 124–127 (Beijing, China, July 25-27 2012)

    2012

  92. [100]

    Transfusion

    Yadan, O. Hydra - A framework for elegantly configuring complex applications, 1.3.2. GitHub (2019). https://github.com/facebookresearch/hydra. 27 Supplementary Note A: Platelet inventory management at UCLH The transfusion laboratory at UCLH maintains a stock of platelets and o...

    2019

  93. [101]

    While our proposed issuing policy could cut wastage with no reduction in service level, a predictive model with a high false positive rate would lead to higher wastage and costs than the baseline OUFO policy. Fig. E1 Contour plots illustrating the performance of our proposed i...

    2015

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.