Article Text
PDF
Abstract
Objective This study identifies preoperative risk factors associated with blood transfusion in women undergoing surgery for ectopic pregnancy. A retrospective chart review of 252 cases was performed at a single academic-affiliated community hospital system between January 2014 and October 2017. Univariate analysis was performed using non-parametric testing where appropriate, and significant variables were incorporated in multivariate modelling.
Methods In this cohort of 252 women, the overall transfusion rate was 8.7% (n=22). Increasing age (OR 1.12, 95% CI 1.03 to 1.22), lower systolic blood pressure (SBP) on presentation (OR 0.96, 95% CI 0.93 to 1.00), lower minimum systolic and diastolic blood pressure (DBP) (OR 0.92, 95% CI 0.89 to 1.95 and OR 0.9, 95% CI 0.84 to 0.93) and lower preoperative haemoglobin (g/L) (OR 0.46, 95% CI 0.33 to 0.62) were associated with higher rates of blood transfusion. Women who had prior care in their current pregnancy and women treated in a women’s specific emergency room (compared with a general emergency room) were less likely to be transfused (OR 0.16, 95% CI 0.05 to 0.51 and OR 0.09, 95% CI 0.03 to 0.30, respectively). Our study used a model in which variance in transfusion can reliably be explained by location of presentation to care, pain alone as a presenting complaint and haemoglobin level (Area under the curve (AUC) =0.87).
Conclusion History of caesarean section and a presenting complaint of pain alone are newly identified indicators for women at higher risk of transfusion at the time of surgical management of ectopic pregnancy. Additionally, care provided in women’s specific emergency rooms may decrease the risk of blood transfusion in this population.
- Family Planning Services
- Gynecology
- Maternal Mortality
- Pregnancy Complications
Data availability statement
Data is available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Blood transfusion is one of the most common complications in women undergoing surgery for ectopic pregnancy, yet risk factors for transfusion in this population have not been well described.
WHAT THIS STUDY ADDS
Lower haemoglobin, abdominopelvic pain and presentation to a general emergency room (vs a women’s specific emergency room) are predictive of transfusion for ectopic pregnancy.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Our study created a model in which variance in transfusion can reliably be explained by location of presentation to care, pain alone as a presenting complaint and haemoglobin level.
Introduction
Ectopic pregnancy occurs at an incidence of 0.64–2.43% of pregnancies.1–4 Management options for ectopic pregnancy include medical treatment with methotrexate injection for women who have an unruptured mass and are haemodynamically stable or surgical treatment by either laparotomy or laparoscopy.5 Roughly 58.6–73.9% of ectopic pregnancies are treated surgically.2 6 7 In 2011–2013, ruptured ectopic pregnancy accounted for 2.7% of all pregnancy-related deaths and was the leading cause of haemorrhage-related mortality.8 Ectopic pregnancy is also associated with significant morbidity in reproductive-age women. Transfusion, life-saving interventions, hysterectomy and sepsis are the most common types of severe morbidity. Despite advances in diagnostic and therapeutic approaches to ectopic pregnancy, severe morbidity may be increasing.4
Rate of transfusion in patients with surgically managed ectopic pregnancies has been quoted as high as 8.6%.9 Transfusion-related complications such as transfusion-associated circulatory overload, fever, allergic reaction, transfusion-related acute lung injury, infection and allosensitisation have the potential to cause long-lasting effects.10 Given these potential risks to young women, transfusion should be a target for prevention.
Identifying women at high risk for transfusion in cases of ectopic pregnancy can be difficult. Current literature is void of guidelines regarding transfusion risk or management in ectopic pregnancy. In a few small studies, haemoglobin <100 g/L, significantly elevated beta human chorionic gonadotropin (ß-hCG) and abnormal uterine bleeding on presentation to care were identified as risk factors for transfusion.9 11
We therefore performed a retrospective cohort study to identify preoperative risk factors associated with perioperative blood transfusion in women undergoing surgery for ectopic pregnancy. Establishing such risk factors provides an opportunity to identify women at risk for transfusion earlier in care, allowing for earlier intervention, modification of protocols and practice, and ideally a reduction in the rate of transfusion.
Methods
This was a retrospective cohort study of women who underwent surgical management of ectopic pregnancy at an academic-affiliated community hospital system between January 2014 and October 2017. Records of 252 patients were reviewed. These records were selected via procedure codes for surgical treatment of ectopic pregnancy (CPT codes are 59120, 59121, 59130, 59135, 59136, 59140, 59150 and 59151). This cohort was then manually reviewed and confirmed by procedure performed as documented in the operative report. We included all women aged >18 years with a preoperative diagnosis of ectopic pregnancy. Both laparoscopic and laparotomy approaches were included. Informed consent was waived for the acquisition of deidentified patient information.
Variables examined include age, body mass index, race and ethnicity history of prior ectopic pregnancy, tubal surgery, pelvic inflammatory disease, caesarean section, diagnosed anaemia or gastrointestinal absorptive disorders, treatment of current pregnancy with methotrexate and prior presentation to care. Prior presentation to care was defined as being seen by a provider (inpatient or outpatient) with a complaint regarding the current ectopic pregnancy. Provider visits unrelated to the ectopic pregnancy were not considered to be a prior presentation to care. Diagnostic criteria such as ultrasound findings, serum ß-hCG levels, haemoglobin on presentation, presenting complaint, location of presentation (general emergency room vs women’s specific emergency room), initial vital signs, and minimum systolic (SBP) and diastolic blood pressure (DBP) (the lowest value recorded during preoperative evaluation) were collected. Surgical details were also assessed, which included time from presentation to the operating room (measured from the time of first recorded vital signs to OR case start time), route of surgery (laparoscopy vs laparotomy), estimated blood loss both inclusive and exclusive of haemoperitoneum, operative time and details of the transfusion.
Descriptive statistics of all variables were performed. Categorical variables were analysed using χ2 or Fisher’s exact test as appropriate. Continuous variables were analysed using the Mann-Whitney U test and Student’s t-test. Univariate logistic regression models were fit for each of the clinically meaningful variables. A p value <0.0500 was considered as significant. Data were analysed by using SAS V.9.4. Haemoglobin on presentation, location of presentation and pain alone as a presenting complaint were used in the multivariate logistic regression model and ORs with 95% CIs were calculated. The performance of the model was evaluated with a receiver operating characteristic curve. STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) case-control reporting guidelines were used.12
Patient and public involvement
Neither the public nor specific patients were involved in the design, conduct or evaluation of this article.
Results
A total of 252 surgeries for ectopic pregnancy were performed between January 2014 and October 2017. The overall transfusion rate was 8.7% (n=22), with the majority (10/22, 45.5%) of those transfused receiving 2 units of packed red blood cells. Patient characteristics are presented in table 1 and were noted to be similar between the two groups, except for increasing age (which was associated with transfusion, mean 33.1 years vs 29.8 years, p=0.0087). History of caesarean section was more common in women who were transfused, while a history of pelvic inflammatory disease was not associated with transfusion. Receiving care in a women’s specific emergency room (compared with presentation to a general emergency room) was protective against transfusion. Women who had received prior care for the pregnancy were less likely to be transfused. No patients who had undergone methotrexate therapy were transfused (table 1).
Characteristics of women transfused and not transfused who underwent surgery for an ectopic pregnancy
Clinical indicators of blood loss associated with higher rates of transfusion included lower SBP on presentation, lower minimum pressures (SBP and DBP) and lower preoperative haemoglobin (table 1). Women who presented with complaints of pain alone (vs vaginal bleeding+/-pain or abnormal diagnostic testing) had higher rates of blood transfusion. Complex-free fluid was seen on ultrasound in 90.5% of women who were transfused, compared with 59.2% of women who were not transfused (p=0.0041). Maternal heart rate, quantitative ß-hCG and ultrasound characteristics of the pregnancy were not associated with transfusion risk.
Perioperative care assessment revealed patients who were transfused arrived in the operating room in less time. This trended towards, but did not reach, statistical significance (287 min vs 385 min, p=0.0803). Laparotomies represented only 8.7% (n=22/252) of total cases, and yet they accounted for 45.5% (10/22) of the transfusions. Thirteen of the 22 laparotomies were conversions from laparoscopy. Documented reasons for conversion included obscured visualisation due to the large haemoperitoneum and cornual location of ectopic pregnancy. Of the 230 cases (91.3% of the total) managed laparoscopically, only 12 (5.2%) were associated with transfusion (table 2).
Perioperative details
ORs for clinically relevant and significant presurgical characteristics which influence transfusion include presentation to a women’s specific emergency room instead of a general emergency room (OR 0.09, 95% CI 0.03 to 0.29) and an increased haemoglobin (g/L) on presentation (OR 0.46, 95% CI 0.33 to 0.62), which are protective, and history of caesarean section (OR 3.73, 95% CI 1.36 to 10.21) and pain alone as a presenting complaint (OR 4.20, 95% CI 1.55 to 11.57), which are adverse (table 3). A multivariate logistic regression model demonstrates that variance in transfusion can reliably be explained by location of presentation to care, pain alone as a presenting complaint and haemoglobin level (AUC=0.87) (figure 1, table 4).
Multivariate logistic regression model. Receiver operating characteristic (ROC) curve for blood transfusion in patients undergoing surgical management of ectopic pregnancy in relation to preoperative haemoglobin (g/L), presenting complaint and location of presentation.
ORs for transfusion in clinically significant variables
Performance of the model was evaluated with a receiver operating characteristic curve
Discussion
In this study of women undergoing surgical management of ectopic pregnancy, key indicators of blood loss (including lower SBP and DBP and lower preoperative haemoglobin) are confirmed to be risk factors for blood transfusion, and novel risk factors, including history of caesarean section, a chief complaint of pain alone and presentation to the general emergency room (compared with a women’s specific emergency room) are identified. Laparoscopic surgical management was associated with significantly lower rates of transfusion compared with laparotomy. Maternal heart rate, quantitative ß-hCG, ectopic size (by ultrasound), anatomic location of the ectopic pregnancy, history of pelvic inflammatory disease and intraoperative lysis of adhesions were not associated with transfusion risk.
Previous studies, of which there were two on review of the literature, identified haemoglobin <100 g/L, ß-hCG>=6500 mIU/mL and abnormal uterine bleeding on presentation to care as risk factors for transfusion in cases of ectopic pregnancy.9 11 Our study mirrored some of these findings such as an increased risk of transfusion with a lower haemoglobin level; conversely, we found no association between ß-hCG level and transfusion risk. Furthermore, newly identified related factors were elucidated within our study. These include a history of caesarean section and a chief complaint of pain at presentation as risk factors for transfusion, while presentation to a women’s specific emergency room was found to be protective.
Interestingly, traditional teaching highlights tachycardia as an early sign during the compensatory phase of hypovolaemic shock which may signal providers to initiate a blood transfusion as one of the resuscitative measures.13 14 On the contrary, in our study, elevated heart rate was not a risk factor for transfusion in this cohort, and instead a lower presenting SBP and/or DBP value was associated with increased risk of transfusion.
Notably, no patients who had undergone methotrexate therapy were transfused. Although this difference did not reach statistical significance (p=0.0518), it suggests that methotrexate therapy had been administered and managed appropriately. Likewise, patients who had been followed up as outpatients and those with prior presentation to care were less likely to be transfused (OR 0.16, 95% CI 0.05 to 0.51) suggesting that delayed surgical management for these women was safe and did not increase their risk of transfusion.
We confirmed that key indicators of blood loss, including SBP, DBP and initial haemoglobin value, correlate with the risk of transfusion. Novel findings that women presenting with a chief complaint of pain alone and women with a history of caesarean section are also at higher risk of transfusion suggesting these women may require more active management to prevent transfusion. Providers and protocols should be updated to account for these findings.
Women managed surgically with laparotomy were at much higher risk of transfusion (45.5%) compared with women who underwent laparoscopic surgery (5.2% transfusion risk) in our study. While 13 of the 22 laparotomies were conversions from laparoscopy, indications for conversion included obscured visualisation due to large haemoperitoneum and cornual location of ectopic. Prior studies have shown that laparoscopic management of ectopic pregnancies, including those that are ruptured and even those in resource poor settings, is feasible and safe and is associated with lower haemoperitoneum volume, shorter operating room time and significantly less morbidity.15–17 While no formal recommendation currently exists as to the mode of surgery, laparoscopy is preferred over laparotomy when feasible.18 Findings such as the disproportionate rate of transfusions demonstrated in our study should prompt gynaecological surgeons to re-examine the standard of care and perhaps develop an expectation of laparoscopic management of ectopic pregnancies in even more challenging cases. The observed correlation between laparoscopic surgery and a reduced risk of transfusion may, in part, be attributed to the fact that patients undergoing laparotomy are frequently more unstable, often as a result of significant haemoperitoneum. In cases of patient instability, laparotomy may be the appropriate surgical route. Additionally, in cases where a provider does not possess the skill of laparoscopic suturing, a laparotomy should be performed for cornual ectopic pregnancies. While our study was not powered to specifically evaluate differences in route of surgical management, given the growth of experience and expertise in minimally invasive gynaecological surgery, this subject should be further explored.
The most significant implication of the findings from this study is that women who presented to a women’s specific emergency room had significantly decreased transfusion rates. The difference in care cannot be explained by time alone because the time from arrival to the operating room was more rapid in those that were transfused (286 min vs 385 min, p=0.0803), although not significant. This time difference may be skewed by those patients who arrived unresponsive and/or were taken directly to the operating room due to instability. One possible explanation is that some patients are already known to the women’s specific emergency room via tracking of ß-hCG for patients with pregnancies of unknown locations. Nevertheless, more information is needed to identify potential risks and benefits of women’s specific care units. Emergency rooms and facilities may need to consider developing care units dedicated to women’s emergencies if these results are confirmed on a broader scale.
A novel finding exclusive to our study is that patients’ risk of transfusion during surgical management of ectopic pregnancy drastically increased (RR 8.10, p<0.0001) when they presented to our general emergency room versus our women’s specific emergency room. This finding calls into question whether specialised care and greater familiarity with female disease processes is beneficial to this patient population or if patients presenting to the different locations are dissimilar cohorts. In our hospital, the women’s specific emergency room is staffed by Women’s Health providers only and is adjacent to Labour and Delivery, physically separate from the main emergency department. Currently, the literature is void of studies evaluating women’s specific emergency room care. This should be further investigated at a national level to examine the prevalence of this care and the potential benefit it provides. Additionally, future studies should be performed to identify specific differences in the triage and evaluation of these patients in a women’s versus general emergency room. This would allow for new protocols to be developed to facilitate care in all emergency areas.
Our study unveiled previously unrecognised risk factors for transfusion during surgical management of ectopic pregnancy including the history of caesarean section (RR 3.21, p<0.0105) and a presenting complaint of pain (RR 5.0, p<0.0002)—these should also be studied and validated for generalisability in other populations.
On a national level, surgical conversions from laparoscopy should be studied in this and other gynaecological procedures to identify risk factors for conversion to laparotomy, blood transfusion, infection and other associated morbidity.
This study adds to the literature in an area in which investigation has been sparse. By examining a broad range of patient characteristics, preoperative workup values and surgical details, we identify previously unknown risk factors for transfusion. Most notably, this is the first study to demonstrate potential improvement in care for women through specialised emergency areas. Limitations of this study include a small sample size, a single institution and a multivariate logistic regression model which has not yet been validated in a second population.
Chief complaints of pain alone and a history of caesarean section are newly identified risk factors for blood transfusion at the time of surgical management for ectopic pregnancy which need to be validated in broader populations. Importantly, receipt of care in a women’s specific emergency room was strongly associated with a decreased risk of transfusion as compared with care in a general emergency room. Focused care areas should be further investigated to see if women’s specific care units provide benefit for this and other emergencies.
Data availability statement
Data is available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Approval for this study was obtained by the Christiana Care Institutional Review Board, CCC# 40 041.
Footnotes
Contributors RC contributed to the conception, design, acquisition, analysis, drafting, final approval of this paper and is the guarantor. CJ, TH, MP and GM contributed to the conception, acquisition, drafting and final approval of this paper. YY and MH and SS contributed to the analysis, interpretation of data for the work, drafting and final approval of this paper. All authors agree to be accountable for all aspects of the work.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.