Comparison of heparin assay monitoring protocols (CHAMP): study protocol for a pragmatic randomised trial Authors: Grace Van Winkle, Benjamin Tillman, Leena Choi, Sarah Berardi, Mary L Dear, Ariel Lewis, Todd W Rice Published: 2026 ABSTRACT AbstractIntroductionDue to widespread availability and familiarity, unfractionated heparin (UFH) is the most used intravenous anticoagulant for many indications in hospitalised patients. UFH, however, is a high-risk medication with complex pharmacokinetics and pharmacodynamics that are highly variable between different patients and within the same patients over time. The traditional titration and monitoring approach uses a clot-based assay, the activated partial thromboplastin time (aPTT), titrated to one and a half to two and a half times the upper limit of the normal range. Alternate assays indirectly measuring the anti-Xa level have not been compared with the aPTT for the monitoring of heparin in a prospective study. The optimal laboratory test for monitoring and adjusting heparin is not known. Our pragmatic study developed within the learning healthcare infrastructure is designed to answer this clinical question by comparing two established protocols for monitoring heparin in our hospital through a pragmatic randomised clinical trial.Methods and analysisThe Comparison of Heparin Assay Monitoring Protocols (CHAMP) Trial is a single-centre, pragmatic, randomised trial conducted at Vanderbilt University Medical Centre (VUMC) beginning 26 June 2024. The CHAMP trial compares the aPTT protocol to the anti-Xa protocol for monitoring and titration of intravenous UFH for systemic anticoagulation in hospitalised adult patients. Admitted patients initiated on UFH protocols are assigned to either the aPTT or anti-Xa protocol in a randomised fashion. The primary outcome is time to reach the therapeutic anticoagulation range by coagulation assay. Secondary outcomes include the percent of measurements in the therapeutic range, the number of coagulation laboratory measurements over time, frequency of heparin rate changes while on the protocol and the incidence of thrombotic and clinically relevant bleeding events.Ethics, waiver of informed consent and disseminationThe CHAMP trial is an ongoing pragmatic trial embedded into the current existing clinical workflow for heparin administration. The two protocols are considered clinically equivalent and already used in clinical practice, allowing for a waiver of consent approval (VUMC Institutional Review Board #232192). This waiver is critical to the implementation of the study due to the nature of scenarios and time constraints in which UFH is typically initiated. Partnering with nursing, pharmacy and clinical providers has been key to launching this study, which provides the first prospective, randomised, direct comparison of the two heparin laboratory protocols available for the monitoring and titration of intravenous UFH in hospitalised patients. After trial completion and data analysis, the findings of the CHAMP trial will be submitted to a peer-reviewed journal for consideration of publication for distribution to a broad clinical audience.Trial registration numberThe CHAMP study was registered on ClinicalTrials.gov (NCT identifier: NCT06329921) on 19 March 2024. The first patient was enrolled in the study on 26 June 2024, with enrolment of the planned 700 participants expected to occur over two years.Keywords: Randomized Controlled Trial, Anticoagulation, Pragmatic Clinical Trial STRENGTHS AND LIMITATIONS OF THIS STUDY This is a pragmatic and randomised study across a broad range of hospitalised adults. This is a pragmatic and randomised study across a broad range of hospitalised adults. This is the only study comparing the activated partial thromboplastin time (aPTT) and anti-Xa protocols in a prospective fashion. This is the only study comparing the activated partial thromboplastin time (aPTT) and anti-Xa protocols in a prospective fashion. There are a significant number of patients on chronic anticoagulation with anti-Xa inhibitors which has resulted in exclusion by laboratory criteria for a higher than anticipated number of patients. There are a significant number of patients on chronic anticoagulation with anti-Xa inhibitors which has resulted in exclusion by laboratory criteria for a higher than anticipated number of patients. The two anticoagulation protocols are algorithm-based but the aPTT protocol is managed by nursing staff and the anti-Xa protocol is managed by pharmacy. Though this represents real-world conditions, it is a confounding variable. The two anticoagulation protocols are algorithm-based but the aPTT protocol is managed by nursing staff and the anti-Xa protocol is managed by pharmacy. Though this represents real-world conditions, it is a confounding variable. The anti-Xa assay is not as widely available at some institutions. The anti-Xa assay is not as widely available at some institutions. Introduction Unfractionated heparin (UFH) is the most frequently used intravenous anticoagulant due to its widespread affordability, availability, rapid onset of action and reversibility with an available antidote. It is most often employed for treatment and prevention of thrombotic conditions including acute coronary syndrome, ischaemic stroke and venous thromboembolism (VTE). There is significant interpatient variability in response to UFH, and the potential for unpredictable pharmacokinetics results in a narrow therapeutic index. As a result, UFH is restricted to the inpatient setting where close monitoring is possible.1 Parenterally administered UFH has an immediate onset of its anticoagulant effects. The activated partial thromboplastin time (aPTT) is the traditional clot-based assay used for estimating the degree of anticoagulation and adjustments. Monitoring heparin with the aPTT involves secondary titration, with prolongation of the clotting assay representing a surrogate marker of the desired primary physiologic outcome: inhibition of coagulation in vivo.2 The aPTT assay requires individual laboratory calibration to a specific heparin concentration to account for the variable aPTT responses between labs due to different reagents or instruments. The institution or lab target aPTT is calibrated to a therapeutic heparin concentration (0.3–0.7 U/mL).3 As a global assessment of coagulation, the aPTT depends on both the intrinsic and common pathways of the coagulation cascade, which is influenced by many factors. The pre-analytic contributions (sample processing), analytic effects (laboratory instruments and reagents) and patient variables (increased endogenous levels of clotting factors or heparin binding proteins) all contribute to interpatient variability in results. Certain clinical scenarios that can result in a prolongation of the aPTT that is not representative of an anticoagulant effect, such as a lupus anticoagulant, can pose additional challenges in the use of this assay for heparin management.4 The use of the aPTT assay in these conditions may result in overestimations of the anticoagulant effect and potential underdosing of heparin. The opposite scenario (excess heparin administration) may occur if the aPTT fails to appropriately prolong in response to increasing doses of heparin in settings of possible heparin resistance.5 The possibility of false reassurance or therapeutic misguidance from the aPTT assay highlights the value of alternate heparin monitoring strategies. An alternative assay (anti-Xa assay) is available in this setting.6 7 The anti-Xa assay is a chromogenic measurement of the degree of inhibition of activated factor X, which provides a more specific reflection of plasma heparin concentration. While some of the same monitoring limiting factors for the aPTT also apply to the anti-Xa assay, the interpatient physiologic contributions are significantly reduced in comparison with the aPTT assay.3 The anti-Xa assay is considered a clinical equivalent for heparin monitoring and titration, but it is much less known and used among healthcare providers. The anti-Xa protocol is more specific and sensitive than the aPTT algorithm.8 9 Prospective studies of heparin monitoring and titration are needed to determine the most reliable laboratory evaluation. The increasing complexity and number of critically ill patients requiring intravenous heparin who are also at risk for aPTT confounding leaves providers facing uncertainty about the reliability of aPTT monitoring for heparin. Use of the aPTT versus anti-Xa protocols has not been compared in a prospective, randomised fashion.8–10 High-quality comparative effectiveness evidence is needed to inform optimal monitoring of intravenous UFH for systemic anticoagulation in hospitalised patients. In this study, we use the Learning Health System11 infrastructure and health information technology to evaluate these two assays in this ongoing randomised, pragmatic clinical trial at a single centre. Methods and analysis This manuscript was written in accordance with the Standard Protocol Items based on Recommendations for Intervention Trials (SPIRIT) guidelines (table 1, online supplemental table S1).12 The Vanderbilt Centre for Learning Healthcare at Vanderbilt University Medical Centre (VUMC) conducts research studies using a unique model that leverages pragmatic, randomised, controlled clinical trials embedded within usual care.11 The Centre is composed of stakeholders from across the healthcare enterprise including patient representation and supports a wide array of projects evaluating comparative effectiveness, implementation science and pragmatic evaluation approaches. Table 1Participant timeline: schedule of enrolment, interventions and assessmentsStudy periodEnrolmentRandomisationPost-randomisationFinal outcome assessmentTIMEPOINTUFH order entryBaseline labs resulted (up to 5 hours after UFH order)Discontinuation of UFH protocol48 hours after discontinuation of UFH protocolENROLMENT:XEligibility screenXAutomated randomisationXINTERVENTIONS:aPTT protocolXXanti-Xa protocolXXASSESSMENTS:Baseline variablesXPrimary outcomeXSecondary outcomesXSafety outcomesXXExploratory outcomesXXFidelity measurementsXaPTT, activated partial thromboplastin time; UFH, unfractionated heparin. Participant timeline: schedule of enrolment, interventions and assessments aPTT, activated partial thromboplastin time; UFH, unfractionated heparin. Study design The CHAMP trial is a prospective, single-centre, pragmatic, unblinded, randomised and controlled clinical trial being conducted in adult inpatient units at Vanderbilt University Hospital. The trial compares the effectiveness of the aPTT and anti-Xa laboratory protocols for optimal monitoring and titration of intravenous UFH for systemic anticoagulation in hospitalised patients. The primary outcome is time to first reach therapeutic anticoagulation range. The protocol was approved by the VUMC Institutional Review Board (IRB) and registered with ClinicalTrials.gov prior to enrolment starting (NCT06329921). The trial will conduct a single interim analysis with pre-specified efficacy stopping rules, and an independent data and safety monitoring board (DSMB) monitors trial progress and enrolment. Study population Adult inpatients at Vanderbilt University Hospital aged 18 years and older are screened for eligibility if the treating team initiates administration of intravenous UFH using the standard aPTT protocol adjusted by nursing staff. The study is limited to patients on heparin protocols developed for acute coronary syndrome (low dose protocol) or VTE (high dose protocol). Patients receiving intravenous heparin for neurologic indications (acute stroke) or extracorporeal membrane oxygenation are not eligible. Laboratory requirements include a baseline aPTT of ≤36.0 s (hospital reference range 24.0–33.5 s), and a baseline anti-Xa value ≤0.3 units/mL (reference range ≤0.3 units/mL). Patients may also be excluded if their provider determines they are not appropriate for the study (table 2). Table 2Inclusion and exclusion criteriaInclusion criteriaExclusion criteriaAge 18 years or olderIndication for anticoagulation is extracorporeal membrane oxygenationAdmitted as observation or inpatient at Vanderbilt University HospitalIndication for anticoagulation is cerebrovascular ischaemic eventOrder placed for unfractionated heparin monitored via the aPTT protocolProvider determines patient is not appropriate for the studyBaseline aPTT assay value is ≥0 and ≤36.0 sBaseline anti-Xa assay value is ≥0 and ≤0.3aPTT, activated partial thromboplastin time. Inclusion and exclusion criteria aPTT, activated partial thromboplastin time. Screening and enrolment Study screening is automated through the EHR to ensure that only patients for whom either protocol would be appropriate are enrolled and randomised. Though both the aPTT and anti-Xa protocols are available and employed in usual care, the nursing managed aPTT is considered routine and the most frequently used protocol in our facility. As the anti-Xa protocol is only requested in specific routine clinical scenarios when there are concerns for aPTT confounding, provider-initiated anti-Xa protocols are excluded from study screening. When a provider initiates an order for intravenous UFH managed via the nurse-managed aPTT protocol for a patient who is 18 years or older, a clinical decision support (CDS) tool alerts the provider of the patient eligibility for the study and queries the provider regarding the presence of any exclusion criteria (figure 1). If none are present, the CDS tool will prompt the ordering provider to submit an order for a single anti-Xa level laboratory order as this result is required for eligibility considerations. If the baseline coagulation laboratory values fall within the specified inclusion range, indicating that the patient is appropriate for monitoring by either protocol, the patient is enrolled and randomised in a 1:1 fashion to either the aPTT or anti-Xa monitoring protocol (figure 2). Provider-facing clinical decision support alert. PTT, partial thromboplastin time. Schematic procedure of arm assignment and enrolment. CDS, clinical decision support; PTT, partial thromboplastin time; UFH, unfractionated heparin. Some patients may be discharged, readmitted and receive UFH in a subsequent admission during the study period. In these instances, if the patient meets inclusion criteria and does not meet exclusion criteria during the second admission, the patient will be randomised again, and the readmission will be considered a new instance/encounter. Patients whose heparin is discontinued and re-initiated within the same encounter will not be re-assessed for eligibility or enrolled and randomised again. Study procedures Both the aPTT and anti-Xa monitoring protocols (online supplemental appendices A and B) are employed in this study exactly as they are used in clinical practice. In this medical centre, adjustments to heparin on the aPTT protocol are managed by bedside registered nurses, and the anti-Xa protocol is managed by clinical pharmacists. As a pragmatic investigation, the study is embedded within the usual clinical workflow with broad eligibility criteria to maximise generalisability. Randomisation and treatment allocation Eligible patients are randomised in a 1:1 ratio to the aPTT guided arm or the anti-Xa guided arm. Treatment arm assignments are generated by computerised randomisation in a 1:1 ratio using an EHR based randomization tool. As patient characteristics can necessitate either a high or low dose heparin protocol, randomisation will be stratified by initial order for high and low dose protocols with randomisation to the study arm performed within each stratum. Randomisation will occur only once at the encounter level. Study group assignments will be concealed until it is confirmed that the patient meets all inclusion criteria and no exclusion criteria and the patient has been enrolled. Intervention delivery Patients randomised to the aPTT arm will remain on this protocol as originally ordered by the provider. For patients randomised to the anti-Xa protocol, the CDS tool will send an automated message to the pharmacist assigned to the primary managing team alerting them that the patient has been assigned to the pharmacy managed anti-Xa protocol. The primary pharmacist for the team will discontinue the nursing managed aPTT protocol and order the pharmacy consult for management by the anti-Xa protocol, as is the standard clinical process for initiating the anti-Xa protocol at Vanderbilt University Hospital. There are no additional prompts or CDS alerts if the UFH is discontinued as it is frequently ordered and discontinued as part of usual clinical care. Blinding As this study relies on using two different heparin management protocols, which both must be known by a provider to be properly executed, it is not possible to blind nursing or pharmacy providers in this study. Additionally, patients will not be explicitly blinded, though it is not common in general practice to discuss the method of monitoring heparin with patients. Since management of the protocols is largely via bedside nursing or pharmacy adjustments, other clinical providers may not be aware of the randomisation group, although it is not hidden from providers in the medical record. Data collection All data are collected as part of routine clinical care. The data are extracted using queries of the electronic health record and stored in a secure online database. Operational data routinely used to monitor heparin protocol compliance will be leveraged to inform intervention fidelity. Variables collected Collected at baseline: age, sex, patient-reported race and ethnicity, body mass index (BMI), sequential organ failure assessment (SOFA) score, active comorbidities (Elixhauser Comorbidity Index), active liver disease, most recent haemoglobin, most recent platelet count and the heparin target dose (eg, high dose or low dose) ordered at baseline. Collected from the time of initiation of the UFH infusion to 48 hours after the heparin protocol is discontinued: date/time of heparin protocol initiation and discontinuation during encounter, coagulation lab values, UFH rate changes, thrombotic events, haemoglobin, transfusions of red blood cells, operational and lab metrics describing protocol compliance. Data monitoring Given the minimal risk of the study procedures, the Principal Investigator (PI) is responsible for overseeing the safety of this trial on a daily basis. The PI will regularly monitor trial data, assess trial operations and record and report adverse events to the IRB in accordance with local regulations. DSMB An independent DSMB composed of experts in haematology, pharmacy, biostatistics and clinical trials will monitor the trial progress. As this study is expected to pose minimal risk, the principal role of the DSMB is to monitor the data and integrity of the trial. Primary outcome The primary outcome is time to reach the first therapeutic anticoagulation range by coagulation assay. By definition, every patient begins out of therapeutic range. As UFH takes effect rapidly,13 it can be presumed that patients found to be in therapeutic range became so shortly after their heparin was administered or adjusted. Patients found to be in therapeutic range at the first lab measurement will be considered to reach therapeutic range 10 min after the heparin infusion initiation. Patients found to be in therapeutic range at subsequent measurements will be considered in range 10 min after the last heparin infusion rate adjustment. Secondary outcomes Secondary effectiveness outcomes include: (1) Percent of laboratory measurements (aPTT or anti-Xa) within the therapeutic range, as defined by protocol, (2) The total number of coagulation laboratory measurements while receiving UFH for overall in-hospital coagulation time (ie, while receiving intravenous UFH infusion) and (3) The total number of heparin rate changes for overall in-hospital coagulation time. Secondary safety outcomes include the incidence of the following outcomes: New thrombotic events on anticoagulation or within 24 hours of anticoagulation cessation. New thrombotic events on anticoagulation or within 24 hours of anticoagulation cessation. Any clinically relevant bleeding adverse events, defined as:Fatal bleeding.Overt bleeding events causing a decline in haemoglobin >2 g/L over a 24-hour period.Bleeding leading to transfusion of two or more units of whole blood or red blood cells, within 48 hours of anticoagulation cessation.Intracranial bleeding events. Any clinically relevant bleeding adverse events, defined as: Fatal bleeding. Fatal bleeding. Overt bleeding events causing a decline in haemoglobin >2 g/L over a 24-hour period. Overt bleeding events causing a decline in haemoglobin >2 g/L over a 24-hour period. Bleeding leading to transfusion of two or more units of whole blood or red blood cells, within 48 hours of anticoagulation cessation. Bleeding leading to transfusion of two or more units of whole blood or red blood cells, within 48 hours of anticoagulation cessation. Intracranial bleeding events. Intracranial bleeding events. Any new thrombotic events on anticoagulation or any clinically relevant bleeding adverse events as defined above. Any new thrombotic events on anticoagulation or any clinically relevant bleeding adverse events as defined above. Exploratory outcomes Time on heparin drip protocol out of therapeutic range for overall in-hospital coagulation time. As with the primary outcome, all patients will begin out of therapeutic range starting at the time of heparin infusion initiation. Patients will be considered out of therapeutic range until a lab result shows they are within range, and vice versa. Patients whose first lab measurement is within therapeutic range will be considered in range starting 10 min after heparin infusion initiation. Patients who were previously out of range and become in range following an infusion rate adjustment will be considered in range beginning 10 min after the most recent rate change. Patients with a lab measurement out of therapeutic range after previously being in range will be considered out of range from the time the lab specimen was taken. Time on heparin drip protocol out of therapeutic range for overall in-hospital coagulation time. As with the primary outcome, all patients will begin out of therapeutic range starting at the time of heparin infusion initiation. Patients will be considered out of therapeutic range until a lab result shows they are within range, and vice versa. Patients whose first lab measurement is within therapeutic range will be considered in range starting 10 min after heparin infusion initiation. Patients who were previously out of range and become in range following an infusion rate adjustment will be considered in range beginning 10 min after the most recent rate change. Patients with a lab measurement out of therapeutic range after previously being in range will be considered out of range from the time the lab specimen was taken. Percentage of patients who reach therapeutic range by first measurement, second measurement and within 24 hours after initiation. Each of these is a binary outcome (Yes/No), assessed cumulatively. Percentage of patients who reach therapeutic range by first measurement, second measurement and within 24 hours after initiation. Each of these is a binary outcome (Yes/No), assessed cumulatively. UFH dosing (units/kg/hour and units/hour) at the time of reaching therapeutic range, for patients who reach therapeutic range. UFH dosing (units/kg/hour and units/hour) at the time of reaching therapeutic range, for patients who reach therapeutic range. Protocol compliance (Yes/No). Protocol compliance definitions are provided in online supplemental 2, appendix A. Protocol compliance (Yes/No). Protocol compliance definitions are provided in online supplemental 2, appendix A. Adjusted scale of anticoagulation time measurements: the following measures will be evaluated using an adjusted scale for anti-Xa and aPTT values, designed to align their ranges for comparability. On this scale, a value of zero (the midpoint of the therapeutic range) represents the optimal clinical status, with deviations in either direction indicating worsening conditions. Details of the adjusted scale are provided in online supplemental 2 appendix A, online supplemental figures S2– S4.Root mean squared difference from zero on the adjusted scale.Mean absolute difference from zero on the adjusted scale. Adjusted scale of anticoagulation time measurements: the following measures will be evaluated using an adjusted scale for anti-Xa and aPTT values, designed to align their ranges for comparability. On this scale, a value of zero (the midpoint of the therapeutic range) represents the optimal clinical status, with deviations in either direction indicating worsening conditions. Details of the adjusted scale are provided in online supplemental 2 appendix A, online supplemental figures S2– S4. Root mean squared difference from zero on the adjusted scale. Root mean squared difference from zero on the adjusted scale. Mean absolute difference from zero on the adjusted scale. Mean absolute difference from zero on the adjusted scale. Fidelity measurements Overall heparin infusion protocol compliance during the first week on heparin drip protocol. The metrics for intervention fidelity are readily available as laboratory measurements in the electronic health record and are tracked by the anticoagulation oversight committee through a Quality, Safety and Risk Prevention dashboard in Tableau. These metrics include protocol compliance to bolus order, lab drawn within time window, infusion rate change, repeat labs and stopping the infusion when assay measurements return far out of range. Overall protocol compliance will be defined as meeting all compliance criteria during this time period. Overall heparin infusion protocol compliance during the first week on heparin drip protocol. The metrics for intervention fidelity are readily available as laboratory measurements in the electronic health record and are tracked by the anticoagulation oversight committee through a Quality, Safety and Risk Prevention dashboard in Tableau. These metrics include protocol compliance to bolus order, lab drawn within time window, infusion rate change, repeat labs and stopping the infusion when assay measurements return far out of range. Overall protocol compliance will be defined as meeting all compliance criteria during this time period. Crossover between protocols will be captured. Crossover between protocols will be captured. Bayesian sequential clinical trial design and interim analysis This study uses a Bayesian sequential clinical trial design to compare two coagulation assay methods (ie, the aPTT guided arm and the anti-Xa guided arm) through simulation studies using preliminary data collected during the 6 months preceding study initiation. The maximum sample size is set at 700 patients over a 24 month study period, based on the estimated enrolment rate of approximately 30 eligible patients per month. The type I error rate and the power to detect superiority between study arms across a range of true HRs are shown in figure 3 and online supplemental figure S1. Operating characteristics of Bayesian sequential clinical trial design. The red line denotes the type I error rate, and the other curves represent the power to detect superiority of one arm over the other across a range of true HRs. Interim analysis One interim analysis will be conducted in a blinded manner within the Bayesian framework after 500 patients have been enrolled and their primary outcome data have been completely collected. At this interim, a Bayesian survival analysis for the primary outcome will be performed. A sceptical prior will be placed on log HR, the parameter of interest for comparing the two arms, specified as a normal distribution with mean 0 and SD 0.35. This prior reflects the belief that extreme effects are unlikely, corresponding to a 5% probability that HR >2 or HR <0.5. We will also calculate the posterior predictive probability of success at the final analysis given the interim data, where success is determined if the predictive probability that HR <1 exceeds 0.975 or the predictive probability that HR >1 exceeds 0.975 at the planned maximum sample size. At the interim analysis, two rules will be applied: Efficacy stopping rule (binding): the trial will be stopped early for superiority of one arm over the other if the posterior probability that HR >1 exceeds 0.975 or the posterior probability that HR <1 exceeds 0.975. Efficacy stopping rule (binding): the trial will be stopped early for superiority of one arm over the other if the posterior probability that HR >1 exceeds 0.975 or the posterior probability that HR <1 exceeds 0.975. Futility decision rule (non-binding): a decision to enrol the remaining patients up to the sample size of 700 will be guided by the posterior predictive probability of success. If this probability is <0.20 at the planned maximum sample size, the trial will be considered futile. Because this rule is non-binding, the final decision at the interim will take into account the totality of evidence. Futility decision rule (non-binding): a decision to enrol the remaining patients up to the sample size of 700 will be guided by the posterior predictive probability of success. If this probability is <0.20 at the planned maximum sample size, the trial will be considered futile. Because this rule is non-binding, the final decision at the interim will take into account the totality of evidence. If the trial is not stopped at the interim, enrolment will continue until the maximum sample size of 700 patients is reached. Statistical analysis plan The full statistical analysis plan (SAP V.2.0, finalised 7 April 2026) is available in online supplemental file 2. Patient baseline demographic and clinical data will be described overall and by the study arm using descriptive statistics. Descriptive analysis of safety, exploratory and fidelity outcomes will be performed by study arm as well as low versus high dose heparin protocol. The full analysis set will include all randomised participants who are eligible and did not meet exclusion criteria specified in the study protocol. All statistical analyses will be performed using the full analysis set with the study arms as originally assigned, even if participants randomised to anti-Xa guided arm never initiate the anti-Xa protocol (ie, these participants remain on the aPTT protocol) or if the participants in the aPTT protocol get switched to anti-Xa monitoring for clinical purposes. Main analysis of the primary, secondary and exploratory outcomes The interim and main analyses for the primary outcome will be conducted using a Bayesian approach, while analyses of all other outcomes will be performed within a frequentist framework. Primary outcome A survival analysis using a Bayesian approach will be conducted to assess whether the time to reach the therapeutic anticoagulation range differs between the aPTT guided and anti-Xa guided arms. The analysis will adjust for the following covariates: age, sex, race, ethnicity, BMI, baseline SOFA score, Elixhauser comorbidity score, duration of heparin protocol (≥24 hours vs <24 hours) and baseline heparin target dose (high vs low). A sceptical prior (as defined in the Interim Analysis section) will be placed on the parameter for the treatment arm assignment, while weakly informative priors will be used for all other parameters. Secondary outcomes Analyses of the secondary effectiveness outcomes will be conducted using generalised linear models with appropriate link functions, within a frequentist framework, adjusting for the same covariates listed above. For analyses of the number of coagulation laboratory measurements and the number of heparin rate changes, overall in-hospital coagulation time will be included as an offset term to account for differences in duration of observation. Descriptive analysis of secondary safety endpoints will be performed by study arm as well as low versus high dose heparin protocol. Handling intercurrent events The death rate is expected to be very low. If death occurs, it will be handled using a while on treatment strategy. Handling of missing data We do not expect missingness in the primary and secondary outcomes. For missing covariates, if the missingness occurs in <5% of the sample, a single imputation will be used to impute missing values; otherwise, multiple imputation will be used for missing covariates. Differential treatment effects To assess whether the effect of two coagulation assay methods on outcomes is modified by patient characteristics, the models for the primary and secondary outcomes will include interaction terms between study arm assignment and the following covariates: baseline heparin target dose (low- vs high-dose), duration of heparin protocol (≥24 hours vs <24 hours), and baseline SOFA score. A sensitivity analysis will be conducted for SOFA score to evaluate potential nonlinearity. Unless there is evidence for nonlinearity, the model will assume a linear interaction between study arm assignment and SOFA score. Subgroup analysis Subgroup analyses of the primary and secondary effectiveness outcomes will be conducted for patients on low- versus high-dose heparin protocols, as well as for patients with protocol duration ≥24 hours versus <24 hours, if sufficient evidence supports differential treatment effects in these groups. Patient and public involvement Patients or the public were not involved in the design, conduct, reporting or dissemination plans of the research; however, the Vanderbilt Centre for Learning Healthcare has public and community representation on the LHS Steering Committee, which contributes to the design of all LHS-supported research projects. Ethics, waiver of informed consent and dissemination Waiver of informed consent This study compares two standard-of-care protocols currently being used in clinical practice at Vanderbilt University Hospital to manage UFH in hospitalised adults. Both the aPTT and anti-Xa protocols are considered clinically acceptable methods for anticoagulation monitoring both at this hospital and nationally. UFH is administered to patients in whom an immediate onset of anticoagulation is desired. The clinical scenarios include patients at high risk for complications or adverse outcomes if care is delayed. As patients would receive an accepted standard of care in either arm, the research risks are not greater than those incurred in routine clinical care. Additionally, as providers may choose to exclude their patient or change the heparin monitoring assay used based on clinical indications, the risks of the research would not be different from what patients would experience as part of their standard clinical care and provider autonomy is preserved. Obtaining informed consent for participation in this study would be impracticable, as this would delay the initiation and initial rate changes for the heparin infusion being used in the patients’ routine care. UFH is administered emergently after it is ordered; attempting to consent patients for randomisation to a heparin monitoring protocol would result in a detrimental delay of care solely for research purposes, as patients are not consented in clinical practice for its administration or monitoring. Furthermore, obtaining consent would introduce a delay in obtaining the anti-Xa level (being run using the commercially available assay on plasma sample already drawn for clinical purposes) and a potential delay in initiating the anti-Xa monitoring protocol, decreasing the potential benefit of monitoring heparin infusions via anti-Xa protocols. The rights and welfare of participants would not be adversely affected by the study, as these assays and protocols are normally encountered in usual care and participants are not involved in the decision of how to monitor intravenous heparin dosing. Additionally, the pre-specified stopping rules incorporated in this adaptive trial design will help ensure that patient participation yields clinically and scientifically meaningful findings and that no more patients than are necessary to answer the question are enrolled. Because the study is expected to pose minimal risk, would not adversely affect the rights and welfare of patients, and prospective informed consent is considered to be impracticable, a waiver of informed consent was requested and granted by the VUMC IRB (#232192). Protocol changes Any changes to the CHAMP protocol will be documented, and amendments will be posted to ClinicalTrials.gov. Dissemination Upon conclusion, the CHAMP trial will be the first study directly comparing the aPTT and anti-Xa protocols within clinical practice in a prospective and randomised fashion. Randomisation within our study provides more reliable and robust data to inform providers of the best option for heparin titration. After trial completion and data analysis, the findings of the CHAMP trial will be submitted for presentation at local, national and international medical conferences, as well as academic institutions through visiting and invited lectureships. Trial results will be submitted to a peer-reviewed journal for consideration of publication for distribution to a broad clinical audience. Following publication of primary results, deidentified data will be made available on reasonable request to the corresponding author in accordance with institutional regulations. Discussion The increasing complexity and number of hospitalised patients requiring intravenous heparin who are also at risk for variable aPTT reliability and confounding leaves providers facing uncertainty about the use of aPTT monitoring for heparin. Alternative strategies using anti-Xa monitoring are also used in current clinical care. To determine the most reliable laboratory monitoring approach, prospective studies of heparin monitoring strategies are needed. Reaching therapeutic anticoagulation more quickly and reliably may decrease a patient’s risk of experiencing thrombosis or haemorrhage and also impact labour and workflow considerations for health system personnel. As such, the results of this pragmatic trial will inform optimal strategies for monitoring of UFH in acutely ill hospitalised patients. This trial has many strengths. The pragmatic design, conducted within the clinical care of hospitalised patients, incorporates the use of broad eligibility criteria in real-world setting and population. EHR-based screening for eligible patients and automated messages to providers facilitate the delivery of the assigned intervention. These elements increase the generalisability and external validity of the findings. Additionally, individual patient randomisation balances patient, hospital and clinical team factors and allows for direct comparison of the two monitoring strategies and potential assessment of a causal relationship between monitoring strategy and outcomes. The trial also has limitations. This is a single-centre trial excluding patients with ischaemic stroke or on extracorporeal membrane oxygenation, limiting the generalisability to these groups. Although randomisation is concealed, clinical providers and patients are not blinded to the assigned monitoring strategy. The two monitoring strategies also involve different providers, which limits our ability to directly compare the monitoring labs or the providers. This difference only allows comparison of the combination strategies of the monitoring lab with the managing providers, namely aPTT monitoring via nursing with anti-Xa monitoring by pharmacy. However, this reflects the real-world practice of these heparin monitoring strategies in the study hospital. Finally, the incidence of relevant thrombotic or haemorrhagic events is expected to be quite low, reducing the power to detect clinically meaningful differences in these important outcomes. Given this, the study is powered to detect a difference in time to reach therapeutic range, which correlates with a lower risk of bleeding or recurrent thrombosis. Trial status The CHAMP trial is currently enrolling. Enrollment began 26 June 2024 and is expected to continue for 2 years. Registration details The study was registered with ClinicalTrials.gov (NCT06329921) on 19 March 2024 prior to any patients being enrolled. Any amendments to the protocol will be updated on clinicaltrials.gov as per SPIRIT guidelines (table 1, online supplemental table S1). Supplementary material