Background: Pharmacological haemostasis is frequently considered during total abdominal hysterectomy, yet the comparative perioperative profiles of intravenous tranexamic acid and haemocoagulase remain incompletely characterised.Objective:To compare intraoperative estimated blood loss, 24-hour haemoglobin change, units of packed red blood cell transfused and record any significant perioperative events after intravenous tranexamic acid or haemocoagulase administration.Methods:This non-randomised, alternately allocated comparative study included 80 women undergoing elective total abdominal hysterectomy under spinal anaesthesia (40 per group). Tranexamic acid 1 g or Haemocoagulase 1 IU was administered intravenously. Recorded operative blood loss and preoperative and 24-hour postoperative haemoglobin were compared using Welch tests. Categorical outcomes were assessed using Fisher exact tests, with exploratory adjusted regression for blood loss.Results:Mean recorded intraoperative blood loss was 455.50 ± 111.61 Ml in the group that received tranexamic acid and 651.25 ± 192.70 mL in the group that received haemocoagulase (mean difference −195.75 mL, 95% CI −266.12 to −125.38; p<0.001). Mean haemoglobin reduction was 0.95 ± 0.06 versus 1.77 ± 0.91 g/dL (p<0.001). Transfusion was recorded in 0/40 versus 14/40 patients (p<0.001). The groups differed in age, weight and operative duration. Three clinically confirmed recorded blood-loss values did not equal component-derived calculations.Conclusion:The tranexamic acid group had lower recorded blood loss, a smaller 24-hour haemoglobin decline and fewer transfusions compared to hemocoagulase group. These differences require cautious interpretation because treatment allocation was predictable, the groups were imbalanced and blood-loss estimation was not independently validated.
Blood conservation remains central to perioperative care because substantial haemorrhage can obscure surgical dissection, complicate haemodynamic management and increase exposure to allogeneic blood transfusion. Tranexamic acid is a lysine analogue that limits plasminogen-mediated fibrin breakdown, and broad evidence has associated its use with reductions in surgical bleeding and transfusion requirements.[1]. More recent pooled evidence also indicates that the extent of benefit varies considerably between operations, particularly in abdominal surgery, arguing against a universal treatment effect.[2].
Total abdominal hysterectomy carries a variable bleeding burden determined partly by the underlying uterine pathology and the extent of operative dissection. In a randomised study, Sallam and Shady reported favourable bleeding outcomes with intravenous and topical tranexamic acid compared with placebo.[3]. An Indian comparison of topical and intravenous administration found similar blood-conserving effects for both routes,[4] while a systematic review of randomised trials reported lower measured blood loss and fewer transfusions with prophylactic tranexamic acid than with control treatment.[5]. Arvind and colleagues also demonstrated lower blood loss in benign hysterectomy after prophylactic intravenous tranexamic acid.[6] Yet results are not uniform: a 2026 randomised study of difficult abdominal hysterectomy did not find a significant difference in measured blood loss against placebo.[7].
Haemocoagulase preparations act through enzymatic pathways involved in fibrin formation, rather than by inhibiting fibrinolysis. A multicentre trial investigated hemocoagulase Agkistrodon et al., in abdominal procedures[8] and a randomised investigation in orthognathic surgery assessed intravenous haemocoagulase for haemorrhage control.[9] A systematic review in cardiac surgery likewise reported bleeding-related benefits, although different product formulations and surgical settings limit direct extrapolation.[10]. Indian spinal-surgery research has assessed batroxobin, tranexamic acid and their combination as alternative haemostatic approaches.[11]
Direct comparisons between the agents have yielded heterogeneous findings. Qin and colleagues observed lower total blood loss with tranexamic acid than with haemocoagulase atrox during knee arthroplasty.[12] Kim and Han found no significant difference between drug groups in retrospective orthognathic data,[13] whereas Wang and colleagues reported a different pattern in intertrochanteric fracture fixation.[14] Singh and Sahu also described less bleeding with haemocoagulase in lower-limb orthopaedic surgery.[15] These studies have differing intervention doses, operative sites and measurement methods. The present study therefore examined recorded blood loss, 24-hour haemoglobin changes, transfusion requirements and adverse events following the two intravenously administered hemostatic drugs during elective total abdominal hysterectomy.
Study Design and Setting
A single-centre, two-arm comparative clinical study was undertaken in a tertiary health care institute. The available surgical records covered procedures performed from 12 September 2025 to 13 August 2026. Participants were allocated to treatment by strict alternation in clinical sequence, producing 40 women in each group. The allocation was nonrandomised and predictable. No claim of random-sequence generation or concealment was made.
Study Population and Eligibility
The analysis included 80 women undergoing elective total abdominal hysterectomy under spinal anaesthesia. Recorded age ranged from 42 to 62 years, and all patients belonged to American Society of Anesthesiologists (ASA) physical status I or II. The original protocol specified an age range of 40–75 years and excluded patients with bleeding diathesis, known hypersensitivity to the study drugs, higher ASA status or specified major systemic conditions. Individual screening and exclusion logs were not available. Recorded surgical indications included leiomyoma, menorrhagia, dysfunctional uterine bleeding and endometrial hyperplasia.
Medication and Clinical Measurements
Group T received tranexamic acid 1 g intravenously and Group H received haemocoagulase 1 IU intravenously. Both drugs were specified in the clinical protocol for administration before the operative incision, diluted in normal saline. Routine perioperative variables recorded for analysis were age, body weight, ASA status, indication for hysterectomy, operative duration, preoperative heart rate, blood pressure, oxygen saturation, baseline haemoglobin, 24-hour postoperative haemoglobin and transfusions.
Study Outcomes and Blood-Loss Assessment
The primary outcome was recorded estimated intraoperative blood loss (mL). The clinical worksheet separately captured the number of small gauzes, large gauzes and blood-soaked mops, suction volume and irrigation volume. Its nominal component estimate used 10 mL per small gauze, 30 mL per large gauze and 100 mL per mop, adding suction volume and subtracting irrigation volume. For the primary analysis, author-confirmed recorded operative blood-loss values were retained. Component-based reconstruction was concordant in 77 records and discordant in three; no numerical substitution was made for the clinically confirmed estimates. Secondary outcomes were absolute and percentage reduction in haemoglobin at 24 hours, the occurrence and number of packed red blood cell (PRBC) transfusions, and documented hypotension, bradycardia, nausea, vomiting and allergic reactions. These events were not independently adjudicated as drug-related.
Statistical Analysis
Quantitative data are reported as mean ± standard deviation (SD); categorical data are expressed as n (%). Two-sided Welch unequal-variance t-tests compared group means and generated 95% confidence intervals for mean differences, calculated as tranexamic acid minus haemocoagulase. Fisher exact tests were used for binary clinical events and ASA status, and Pearson chi-square testing was applied to the four-category surgical-indication distribution. Exploratory ordinary least-squares models assessed recorded blood loss after adjustment for age, weight, ASA status, surgical indication and preoperative haemoglobin using heteroscedasticity-robust (HC3) standard errors; operative duration was added in a separate sensitivity model. These post hoc adjusted comparisons were not used to establish causal treatment effects. Two-sided p<0.05 was the nominal significance threshold; multiple-comparison adjustment was not applied.
Ethical Considerations
The study protocol was reviewed by the Institutional Ethics Committee. IEC approval number: No. 532/L/11/12/Ethics/ESICMC&PGIMSR/Estt. Vol.IV/3B/2025-2. Patient confidentiality was maintained throughout the study.
All 80 patient records were included (40 per treatment group). Each participant had a recorded baseline and 24-hour postoperative haemoglobin measurement. The treatment sequence alternated without random assignment. The tranexamic acid group was younger and had lower mean weight and shorter procedures than the haemocoagulase group. Differences in demographic and operative characteristics are shown in Table 1. Surgical-indication counts are compared in Figure 1.
Table 1. Baseline clinical and operative characteristics
| Characteristic | Tranexamic acid (n=40) | Haemocoagulase (n=40) | p-value |
| Age (years) | 49.83 ± 4.44 | 53.48 ± 4.81 | p<0.001 |
| Weight (kg) | 62.50 ± 4.76 | 66.30 ± 5.82 | p=0.002 |
| ASA I | 17 (42.5%) | 10 (25.0%) | p=0.155 |
| ASA II | 23 (57.5%) | 30 (75.0%) | |
| Leiomyoma | 15 (37.5%) | 17 (42.5%) | p=0.489 |
| Menorrhagia | 11 (27.5%) | 12 (30.0%) | |
| Dysfunctional uterine bleeding | 9 (22.5%) | 4 (10.0%) | |
| Endometrial hyperplasia | 5 (12.5%) | 7 (17.5%) | |
| Operation duration (min) | 93.18 ± 9.35 | 104.98 ± 13.72 | p<0.001 |
| Preoperative systolic BP (mmHg) | 118.48 ± 5.86 | 124.78 ± 7.48 | p<0.001 |
The mean recorded intraoperative blood loss was 455.50 ± 111.61 mL following tranexamic acid and 651.25 ± 192.70 mL following haemocoagulase. The mean difference was -195.75 mL (95% CI -266.12 to -125.38; p<0.001). Corresponding baseline haemoglobin means were 11.91 ± 0.37 and 11.76 ± 0.44 g/dL; the respective 24-hour values were 10.96 ± 0.40 and 10.00 ± 1.28 g/dL. Mean within-patient reduction was 0.95 ± 0.06 versus 1.77 ± 0.91 g/dL (p<0.001). These observations are summarised in Table 2, and their distributions are visualised in Figures 2 and 3.
Table 2. Recorded blood loss and haemoglobin outcomes
| Outcome | Tranexamic acid (n=40) | Haemocoagulase (n=40) | p-value |
| Recorded intraoperative EBL (mL) | 455.50 ± 111.61 | 651.25 ± 192.70 | p<0.001 |
| Preoperative Hb (g/dL) | 11.91 ± 0.37 | 11.76 ± 0.44 | p=0.113 |
| 24-hour postoperative Hb (g/dL) | 10.96 ± 0.40 | 10.00 ± 1.28 | p<0.001 |
| Absolute Hb reduction (g/dL) | 0.95 ± 0.06 | 1.77 ± 0.91 | p<0.001 |
| Relative Hb reduction (%) | 7.99 ± 0.65 | 15.24 ± 8.34 | p<0.001 |
| All values are mean ± SD. EBL: estimated blood loss; Hb: haemoglobin. Tests compare recorded outcomes, not independent drug effects | |||
Perioperative PRBC transfusion was documented in none of the women receiving tranexamic acid and in 14 (35.0%) receiving haemocoagulase (two-sided Fisher p<0.001). Ten women received intraoperative PRBC transfusion, seven received postoperative PRBC transfusion and three received transfusion at both time points in the haemocoagulase group, accounting for 17 units in total. The transfusion findings are described in Table 3 and illustrated in Figure 4. Transfusions and the timing of postoperative haemoglobin sampling may affect the interpretation of haemoglobin change.
Table 3. Packed red blood cell transfusion requirements
| Measure | Tranexamic acid (n=40) | Haemocoagulase (n=40) | p-value |
| Any intraoperative PRBC | 0 (0.0%) | 10 (25.0%) | p=0.001 |
| Any postoperative PRBC | 0 (0.0%) | 7 (17.5%) | p=0.012 |
| Any perioperative PRBC | 0 (0.0%) | 14 (35.0%) | p<0.001 |
| Intraoperative PRBC units | 0 | 10 | — |
| Postoperative PRBC units | 0 | 7 | — |
| Total PRBC units | 0 | 17 | — |
| A participant could receive PRBC both intraoperatively and postoperatively. Fisher exact tests compare the numbers of participants, not units | |||
The coded perioperative events are presented in Table 4. Hypotension was recorded in 12 haemocoagulase recipients and none receiving tranexamic acid; nausea and vomiting were recorded in nine and six participants, respectively, in the haemocoagulase group. Three allergic reactions were coded in the haemocoagulase group. No bradycardia was recorded in either arm. These descriptive differences do not establish a drug-specific adverse reaction, particularly under spinal anaesthesia.
Table 4. Recorded perioperative clinical events
| Event | Tranexamic acid (n=40) | Haemocoagulase (n=40) | p-value |
| Hypotension | 0 (0.0%) | 12 (30.0%) | p<0.001 |
| Bradycardia | 0 (0.0%) | 0 (0.0%) | — |
| Nausea | 0 (0.0%) | 9 (22.5%) | p=0.002 |
| Vomiting | 0 (0.0%) | 6 (15.0%) | p=0.026 |
| Allergic reaction | 0 (0.0%) | 3 (7.5%) | p=0.241 |
| Multiple events could occur in one patient. P-values are unadjusted Fisher exact tests. Recorded events are not necessarily medication-attributable | |||
In exploratory regression controlling for age, weight, ASA status, surgical indication and preoperative haemoglobin, haemocoagulase assignment was associated with 107.9 mL higher recorded blood loss (95% CI 72.9–142.8; p<0.001). The estimated association was 45.0 mL (95% CI 18.5–71.5; p=0.001) when operative duration was included. These post hoc models are sensitive to specification and should not be considered causal estimates.
This investigation found lower recorded estimated intraoperative blood loss in women allocated to intravenous tranexamic acid than in those allocated to haemocoagulase during elective abdominal hysterectomy. Haemoglobin decline at 24 hours and the frequency of PRBC transfusion were also lower in the tranexamic acid group. The differences are potentially relevant to clinical blood conservation, but they arose in groups formed by predictable alternation rather than genuine random assignment. Furthermore, women receiving haemocoagulase were older, heavier and underwent longer procedures. These characteristics may reflect different operative complexity and confound comparisons between drugs.
In the hysterectomy literature, Sallam and Shady [3] and the systematic review by Abu-Zaid and colleagues [5] reported that prophylactic tranexamic acid can lessen perioperative blood loss compared with placebo or no treatment. The Indian randomised study by Arvind et al.[6] similarly described reduced blood loss and postoperative intervention after prophylaxis, whereas Mitra et al.[4] reported comparable bleeding outcomes between different tranexamic acid routes. These investigations establish plausible benefits of tranexamic acid in women undergoing hysterectomy, but their controls were not equivalent to intravenous haemocoagulase. The more recent difficult-hysterectomy trial by Jindaluang et al.[7] found no significant intraoperative blood-loss reduction with intravenous tranexamic acid against placebo. This divergence emphasises how operative case mix and measurement may modify estimated drug effects.
The present direction of effect resembles the prospective study of Qin and colleagues,[12] who observed lower calculated blood loss with tranexamic acid than haemocoagulase atrox during primary total knee arthroplasty. That trial, however, included combined intra-articular and intravenous administration and assessed bleeding by a haematocrit-based formula. The clinical and pharmacological setting is therefore quite different. Other studies support neither agent as consistently superior: Kim and Han found no significant blood-loss differences among tranexamic acid, Botropase and combination groups in bimaxillary surgery,[13] whereas Wang and colleagues reported different treatment-related patterns in fracture fixation.[14]. Singh and Sahu described an advantage for haemocoagulase in lower-limb surgery.[15] These heterogeneous findings make direct numerical pooling with an abdominal-hysterectomy cohort inappropriate.
Mechanistic differences offer biological plausibility but not a substitute for clinical comparability. Tranexamic acid restricts plasmin-mediated breakdown of fibrin clot, whereas the activity of haemocoagulase preparations depends on their enzyme composition and effects on fibrin formation. Evidence from the abdominal trial by Wei et al,[8] orthognathic work by Shetty and Sriram,[9] cardiac-surgery synthesis by Yao et al.[10] and Indian spinal-surgery research by Nagabhushan et al.[11] demonstrates that haemocoagulase has been evaluated in several operations. Nevertheless, formulations, doses and co-interventions differ substantially. The pharmacodynamic volunteer study by Shenoy et al[16] also illustrates why anticoagulant and fibrinogen-related effects should not be inferred solely from a visible reduction in bleeding.
An important feature of the present results is the disproportionate concentration of transfusions and coded adverse events in one group. These findings warrant careful clinical audit, but they do not mean that haemocoagulase caused hypotension, nausea or vomiting. Spinal anaesthesia, operative duration, fluid balance, uterine pathology and bleeding itself can all influence these outcomes. The available records do not provide detailed temporal relationships between drug administration and each event, nor do they document systematic thromboembolic surveillance. Similarly, postoperative haemoglobin is affected by transfusion, haemodilution and the timing of sample collection. Consequently, the haemoglobin comparison should be regarded as a pragmatic perioperative outcome rather than an isolated pharmacological effect.
The findings also illustrate the sensitivity of the recorded blood-loss difference to measured clinical covariates. Adjustment for age, weight, ASA status, indication and baseline hemoglobin attenuated the treatment-associated difference, and inclusion of operative duration reduced it further. Duration may partly represent underlying case complexity and may also lie downstream of intraoperative events, making simple adjustment methodologically ambiguous. Because uterine size, surgical technique, surgeon identity and standardized transfusion thresholds were not documented, residual confounding remains likely even where model-based p-values are small.
Several limitations must therefore be recognized. This was a single-centre, relatively small study with open, predictable alternating allocation, without allocation concealment. Important group imbalances were present. Blood loss was estimated using nominal absorption volumes assigned to dressings and a suction-irrigation calculation. Three author-confirmed operative blood-loss values did not reconcile with their separate component entries, making independent verification of measurement precision impossible from the clinical worksheet alone. The analysis retained the confirmed operative totals without substituting calculated values. No blinded central assessment, pre-specified causal analysis, uterine weight or comprehensive postoperative safety follow-up was documented. Finally, the available ethics certificate date postdates the first recorded procedures, and institutional verification of its coverage is necessary. These considerations restrict causal inference and generalisability.
The most defensible interpretation is that lower bleeding-related outcomes were observed among women given tranexamic acid under the clinical conditions represented here. Better-controlled studies in abdominal hysterectomy, with standardised blood-loss measurement and documentation of important operative determinants, are needed before concluding that one intravenous drug is therapeutically superior to the other.
In women undergoing elective total abdominal hysterectomy, alternating allocation to intravenous tranexamic acid was associated with lower recorded intraoperative blood loss, smaller 24-hour haemoglobin reductions and fewer PRBC transfusions than allocation to haemocoagulase. Because allocation was nonrandomised and operative characteristics differed between groups, these findings do not independently demonstrate drug superiority. They support further controlled evaluation with robust outcome measurement.