Incidence and risk factor of blood transfusion after abdominal radical hysterectomy for cervical cancer: a 10-year retrospective study of the US nationwide inpatient sample

Abstract

Background

With the rising prevalence of abdominal radical hysterectomy, the need for perioperative blood transfusion has emerged as a significant clinical challenge. Independent risk factors for blood transfusion during abdominal radical hysterectomy remains limited, and identifying these factors is needed.

Methods

A retrospective analysis of data was performed using the Nationwide Inpatient Sample (NIS), focusing on patients who underwent abdominal radical hysterectomy between 2010 and 2019. Patients were categorized into two groups based on whether they received a blood transfusion. The analysis encompassed various demographic factors, including race, sex, and age, as well as length of stay (LOS), total hospitalization charges, hospital characteristics (admission type, insurance type, bed size, teaching status, geographic location, and hospital region), hospital mortality rates, comorbidities, and perioperative complications. Subsequently, both univariate and multivariate logistic regression analyses were employed to ascertain factors associated with abdominal radical hysterectomy patients requiring blood transfusions.

Results

Blood transfusions occurred in 14.84% of patients between 2010 and 2019, with a downward trend over time. Receiving a transfusion was associated with several negative outcomes, including a longer length of stay, higher total charges, and complications like thrombocytopenia, acute myocardial infarction, pneumonia, and so on. Additionally, patients who received transfusions were more likely to experience postoperative delirium, deep vein thrombosis, and wound infection. Independent risk factors for blood transfusion include Black race, Asian or Pacific Islander race, non-elective surgery, hospitalization in a rural setting, pre-existing medical conditions like coagulopathy, chronic blood loss anemia, deficiency anemia and others. Conversely, patients with private insurance, residing in the West, or Midwest/North Central regions were less likely to require a blood transfusion.

Conclusion

Our study highlights the concern of perioperative blood transfusion in radical hysterectomy, linked to significant complications. Reducing intraoperative blood loss and optimizing care based on patient factors are crucial for improving outcomes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-024-13216-3.

Introduction

Approximately 14,000 individuals received a diagnosis of cervical cancer, resulting in 4,000 fatalities in US each year. Cervical cancer is one of the four most common and deadly cancers for women around the world [1]. For patients diagnosed with early-stage cervical cancer, the standard-of-care treatment approach often involves surgical removal of the uterus (radical hysterectomy) alongside the pelvic lymph nodes. This surgery can be performed traditionally through open surgery or by using minimally invasive techniques [2]. Following concerns raised by a 2018 randomized trial by Ramirez et al. [3]. suggesting a potential decrease in overall survival (OS) and disease-free survival (DFS) rates following minimally invasive surgery (MIS) for certain cancers, there has been a notable shift in practice within medical research centers worldwide. This shift is characterized by a resurgence in the use of open surgery [4–6]. The higher incidence of intraoperative hemorrhage and postoperative complications observed in open abdominal radical hysterectomy (ARH) compared to minimally invasive techniques warrant further investigation into the risk factors associated with perioperative blood transfusion during ARH, as well as other potential surgical complications [7–9].

Although blood transfusions are generally safe especially in developed countries due to strict blood screening, they can have some long-term and near-term complications, including acute or delayed hemolytic transfusion reactions, allergic reactions, alloimmunization, transfusion-related lung injury, and infection [10]. Blood transfusions may also cause slower functional recovery, a longer length of hospital stay, and more postoperative complications. Due to low donation rates, transfusion-transmitted infections, suboptimal management of blood stocks, and misallocation, blood for transfusion is an expensive and limited resource.

A recent study by Yang Shen compared intraoperative blood loss between ARH and MIS [11]. However, no prior studies have evaluated independent risk factors for blood transfusion following ARH. This investigation aimed to identify independent variables associated with an increased likelihood of blood transfusion after ARH.

Materials and methods

Database source

The Nationwide Inpatient Sample (NIS) stands as the most extensive all-payer database in the United States, capturing information on inpatient hospital stays [12]. This invaluable resource, sponsored by the Agency for Healthcare Research and Quality (AHRQ), compiles patient-level data from hospital discharges across participating states within the Healthcare Cost and Utilization Project (HCUP) [13]. The NIS offers a wealth of insights into healthcare utilization by including details such as hospitalization length, diagnoses, patient demographics, treatment received, and associated costs [14].

Data extraction

We analyzed data from the NIS database to conduct a retrospective study on women with invasive cervical cancer who underwent abdominal radical hysterectomy between 2010 and 2019. Cases were identified using specific diagnostic codes: ICD-9-CM 1800/1801/1808/1809 for cervical cancer and ICD-9-CM 6869 for ARH. The equivalent ICD-10-CM codes (C530/C531/C538/C539 for cancer and 0UB90ZZ/0UT90ZZ for ARH) were also employed. Since the NIS database is publicly available, as outlined in the Tri-Council Policy Statement (2010), obtaining institutional review board approval was not required. Notably, some patient characteristic data may have been missing from the database. We identified surgical complications using relevant ICD codes. A total of 3,188 unique patients were included in the final analysis. (Fig. 1)

Fig. 1.

Flowchart depicting the patient selection process based on the inclusion/exclusion criteria

Data analysis

The study cohort was divided into two groups based on whether they received a blood transfusion. Data extraction included complications, underlying comorbidities, demographics, and clinical outcomes. The presence of 29 distinct comorbidities was determined using the Agency for Healthcare Research and Quality Comorbidity Software, version 3.7, with these comorbidities flagged using ICD-10-CM and ICD-9-CM diagnosis codes.

To identify potential risk factors for blood transfusions after ARH this study employed a comprehensive statistical analysis. The Kolmogorov-Smirnov test was used to assess the normality of continuous variables. Continuous variables conforming to a normal distribution underwent Student’s t-tests, while non-parametric data were analyzed with the Mann-Whitney U test. Categorical variables were compared using the Chi-squared test. The Chi-square test was used to assess the association between each categorical variable and the outcome variable. Variables with P < 0.05 were selected into the multivariable analysis model. Additionally, variables have been identified as a risk factor for blood transfusion and socioeconomic factor (income level) were included in the regression models after adjusting for confounders [15] (Table 1). Approximately 10% of the data lacked information on patients’ race and ethnicity. To address these missing data and to mitigate the potential biases, we employed the method of multiple imputation. Complete case analysis was used in multivariate models. Hospital-level variables that were significantly associated with the occurrence of blood transfusion in the baseline data were used as stratification variables to further analyze transfusion-related risk factors. Given the limited number of rural hospitals, region of hospital was the sole stratification variable employed in the analysis. Possible interactions between covariates were considered as well. In addition, we performed propensity score matching (PSM) for several socioeconomic factors, including teaching status of hospital, location of hospital, type of insurance, region of the hospital and income level, to reduce potential confounding biases in our analysis. A 2-sided P < 0.05 was considered statistically significant. Statistical analyses were performed using the SPSS Statistics 26.0.

Table 1.

Variables used in binary logistic regression analysis

Variable categories	Specific variables
Patient demographics	Age (≤ 60 years and>60 years), sex (male and female), race (White, Black, Hispanic, Asian or Pacific Islander, Native American and Other), income level (Lowest quartile, second quartile, third quartile, highest quartile)
Hospital characteristics	Type of admission (non-elective, elective), teaching status of hospital (nonteaching, teaching), location of hospital (rural, urban), type of insurance (Medicare, Medicaid, private insurance, self-pay, no charge, other), region of the hospital (northeast, Midwest or north central, south, west)
Comorbidities	AIDS, alcohol abuse, deficiency anemia, rheumatoid diseases, chronic blood loss anemia, congestive heart failure, chronic pulmonary disease, coagulopathy, depression, diabetes (uncomplicated), diabetes (with chronic complications), drug abuse, hypertension, hypothyroidism, liver disease, lymphoma, fluid and electrolyte disorders, metastatic cancer, neurological disorders, obesity, paralysis, peripheral vascular disorders, psychoses, pulmonary circulation disorders, renal failure, solid tumor without metastasis, peptic ulcer disease, valvular disease and weight loss

Result

This study analyzed data from the NIS database to investigate blood transfusion rates following abdominal radical hysterectomy for invasive cervical cancer patients. Over a ten-year period (2010–2019), 50,133 patients with invasive cervical cancer were identified. Among these patients, 3,284 underwent ARH surgery. After excluding patients with missing data (96), a total of 3,188 patients were included in the final analysis. Within this group, 473 patients received blood transfusions, resulting in an overall transfusion rate of 14.84%. Figure 2 depicts the annual trends for both ARH procedures and blood transfusion rates over the course of the study period.

Fig. 2.

Annual cases of abdominal radical hysterectomy and incidence of blood transfusion (2010–2019)

Analysis of patient characteristics revealed several factors associated with blood transfusions. Patients who received transfusions were significantly older (P = 0.005) and remained hospitalized for a longer duration (LOS) (P < 0.001), incurring higher total charges (TOTCHG) (P < 0.001) compared to those who did not. Additionally, factors such as Black or Asian/Pacific Islander race (OR 1.525, 95%CI: 1.134–2.051, P = 0.005; OR 2.135, 95%CI: 1.413–3.226, P < 0.001, respectively), non-elective surgery (P < 0.001), and treatment at a rural hospital (P = 0.014) were associated with an increased likelihood of blood transfusion. Conversely, patients with private insurance (P < 0.001) and those residing in the West (OR 0.674, 95%CI: 0.485–0.937, P = 0.019) or Midwest/North Central regions (OR 0.689, 95%CI: 0.485–0.980, P = 0.038) were less likely to require blood transfusions. While the blood transfusion group exhibited a higher mortality rate, the difference did not reach statistical significance (P = 0.220). (Tables 2 and 3)

Table 2.

Patient characteristics and outcomes after abdominal radical hysterectomy (2010–2019)

Characteristics		Transfusion	No Transfusion	P
Total (n = count)		473	2,715
Total incidence		14.84%
Age (median, IQR, years)		48 (40–60)	47 (38–57)	0.005
Age group
	≤ 60	362(76.2%)	2198(81.0%)	0.015
	> 60	113(23.8%)	515(19.0%)
Race
	White	235(49.5%)	1554(57.3%)	< 0.001
	Black	81(17.1%)	320(11.8%)
	Hispanic	82(17.3%)	473(17.4%)
	Asian or Pacific Islander	39(8.2%)	125(4.6%)
	Native American	14(2.9%)	135(5.0%)
	Other	24(5.1%)	106(3.9%)
LOS (median, IQR, days)		5 (4–7)	3 (2–4)	< 0.001
TOTCHG (median, IQR, $)		68,972(44,005-107,917.5)	48,287(33,220 − 73,577)	< 0.001
Type of insurance
	Medicare	86(13.3%)	363(18.0%)	< 0.001
	Medicaid	143(26.7%)	726(30.2%)
	Private insurance	189(50.9%)	1378(39.7%)
	Self-pay	28(4.2%)	115(5.9%)
	No charge	4(1.2%)	33(0.8%)
	Other	25(3.6%)	98(5.3%)
Bed size of hospital
	Small	32(6.7%)	234(8.6%)	0.383
	Medium	102(21.5%)	561(20.7%)
	Large	341(71.8%)	1918(70.7%)
Elective admission		2449(90.3%)	386(81.3%)	< 0.001
Type of hospital (teaching )		2331(85.9%)	409(86.1%)	0.914
Location of hospital (urban)		1(0.2%)	46(1.7%)	0.013
Region of hospital
	Northeast	92(19.4%)	408(15.0%)	0.006
	Midwest or North Central	78(16.4%)	587(21.6%)
	South	202(42.5%)	1059(39.0%)
	West	103(21.7%)	659(24.3%)
Died		2(0.4%)	4(0.1%)	0.204
Income level
	Lowest quartile	132(28.4%)	784(29.5%)	0.617
	Second quartile	107(23.0%)	661(24.9%)
	Third quartile	119(25.6%)	664(25.0%)
	Highest quartile	107(23.0%)	550(20.7%)

LOS: Length of stay, TOTCHE: Total charge

Table 3.

Risk factors associated with blood transfusion after abdominal radical hysterectomy

Variable		Multivariate Logistic Regression
		OR	95% CI	P
Age > 60 years old		1.101	0.815–1.486	0.531
Race
	White	Ref	——	——
	Black	1.525	1.134–2.051	0.005
	Hispanic	1.119	0.834–1.501	0.455
	Asian or Pacific Islander	2.135	1.413–3.226	<0.001
	Native American	0.800	0.442–1.451	0.463
	Other	1.207	0.736–1.98	0.456
Type of insurance
	Medicare	Ref	——	——
	Medicaid	0.881	0.609–1.275	0.502
	Private insurance	0.637	0.448–0.906	0.012
	Self-pay	1.001	0.583–1.719	0.996
	No charge	0.589	0.196–1.77	0.346
	Other	1.197	0.695–2.06	0.516
Income level
	Lowest quartile	Ref	——	——
	Second quartile	1.043	0.785–1.385	0.773
	Third quartile	1.173	0.885–1.555	0.267
	Highest quartile	1.334	0.986–1.805	0.061
Elective admission		0.504	0.383–0.662	<0.001
Teaching hospital		0.520	0.400–0.690	0.623
Region of hospital
	Northeast	Ref	——	——
	Midwest or North Central	0.689	0.485–0.980	0.038
	South	0.920	0.691–1.226	0.570
	West	0.674	0.485–0.937	0.019

OR: Odds ratio, CI: Confidence interval

The following covariates were controlled: age, race, primary payer, admission type, hospital teaching status, hospital region

Preoperative comorbidities significantly influenced blood transfusion rates. Patients with conditions like coagulopathy (OR 4.33, 95%CI: 2.62–7.15, P < 0.001), chronic blood loss anemia (OR 4.29, 95%CI: 2.64–6.99, P < 0.001), and deficiency anemia (OR 3.95, 95%CI: 2.88–5.42, P < 0.001) were at a considerably higher risk. Similarly, other comorbidities such as fluid and electrolyte disorders (OR 1.94, 95%CI: 1.42–2.64, P < 0.001), peripheral vascular disorders (OR 2.72, 95%CI: 1.32–5.63, P = 0.007), and pulmonary circulation disorders (OR 3.33, 95%CI: 1.05–10.57, P = 0.041) were associated with increased transfusion risk. Additionally, statistically significant differences in blood transfusion rates were observed between groups based on the presence of hypertension (P = 0.001), renal failure (P = 0.001), weight loss (P = 0.001), paralysis (P = 0.015), and congestive heart failure (P = 0.023). (Table 4).

Table 4.

Relationship between blood transfusion and preoperative comorbidities

Comorbidities		Univariate Analysis		Multivariate Logistic Regression
		No transfusion	Transfusion	P	OR	95% CI	P
Preoperative comorbidities
	Coagulopathy	45 (1.7%)	37 (7.8%)	< 0.001	4.33	2.62–7.15	<0.001
	Chronic blood loss anemia	47 (1.7%)	33 (7.0%)	< 0.001	4.29	2.64–6.99	<0.001
	Deficiency anemia	129 (4.8%)	83 (17.5%)	< 0.001	3.95	2.88–5.42	<0.001
	Fluid and electrolyte disorders	216 (8.0%)	91 (19.2%)	< 0.001	1.94	1.42–2.64	<0.001
	Peripheral vascular disorders	26 (1.0%)	13 (2.7%)	0.001	2.72	1.32–5.63	0.007
	Pulmonary circulation disorders	9 (0.3%)	7 (1.5%)	0.004	3.33	1.05–10.57	0.041
	Diabetes (with chronic complications)	44 (1.6%)	16 (3.4%)	0.009	1.95	1.01–3.74	0.046
	Renal failure	43 (1.6%)	18 (3.8%)	0.001	1.31	0.68–2.54	0.417
	Weight loss	62 (2.3%)	24 (5.1%)	0.001	1.18	0.68–2.06	0.561
	Hypertension	761 (28.0%)	167 (35.3%)	0.001	1.17	0.92–1.49	0.202
	Paralysis	6 (0.2%)	5 (1.1%)	0.015	2.58	0.68–9.80	0.164
	Congestive heart failure	23 (0.8%)	10 (2.1%)	0.023	0.85	0.33–2.16	0.724
	Acquired immune deficiency syndrome	10 (0.4%)	2 (0.4%)	0.860	1.06	0.21–5.34	0.947
	Chronic pulmonary disease	279 (10.3%)	58 (12.3%)	0.195	1.00	0.72–1.40	0.981
	Depression	282 (10.4%)	47 (9.9%)	0.766	0.93	0.65–1.31	0.664
	Diabetes, uncomplicated	218 (8.0%)	50 (10.6%)	0.066	1.22	0.85–1.76	0.283
	Drug abuse	47 (1.7%)	12 (2.5%)	0.230	1.19	0.59–2.39	0.633
	Hypothyroidism	219 (8.1%)	32 (6.8%)	0.332	0.74	0.49–1.13	0.160
	Liver disease	45 (1.7%)	8 (1.7%)	0.958	0.55	0.23–1.31	0.177
	Rheumatoid arthritis/collagen vascular diseases	32 (1.2%)	5 (1.1%)	0.820	0.84	0.31–2.30	0.737
	Alcohol abuse	23 (0.8%)	9 (1.9%)	0.061	1.80	0.75–4.35	0.191
	Neurological disorders	31 (1.1%)	9 (1.9%)	0.170	1.08	0.47–2.49	0.852
	Obesity	441 (16.2%)	84 (17.8%)	0.412	1.01	0.76–1.34	0.948
	Psychoses	66 (2.4%)	18 (3.8%)	0.085	1.65	0.92–2.98	0.094
	Valvular disease	26 (1.0%)	9 (1.9%)	0.069	1.66	0.68–4.05	0.263

AIDS: Acquired immunodeficiency syndrome, OR: Odds ratio, CI: Confidence interval

Multivariate analysis revealed a significant association between blood transfusion during ARH. Compared to those who did not receive a transfusion, these patients had a 3.25-fold higher likelihood of thrombocytopenia (OR 3.25, 95%CI: 1.78–5.96, P < 0.001) and a 9.13-fold likelihood of AMI (OR 9.13, 95%CI: 2.97–27.94, P < 0.001).

Perioperative blood transfusion was also independently associated pneumonia (OR 3.56, 95%CI: 1.76–7.17, P < 0.001), urinary tract infection (OR 2.05, 95%CI: 1.21–3.47, P = 0.008), postoperative delirium (OR 10.17, 95%CI: 1.78–57.97, P = 0.009), DVT (OR 2.64, 95%CI: 1.03–6.80, P = 0.044), and wound infection (OR 2.62, 95%CI: 1.23–5.62, P = 0.013) (Table 5).

Table 5.

Relationship between blood transfusion and postoperative complications

Complications		Univariate Analysis			Multivariate Logistic Regression
		No transfusion	Transfusion	P	OR	95% CI	P
Medical complications
	Hemorrhage	23 (0.8%)	26 (5.5%)	< 0.001	5.21	2.77–9.80	<0.001
	Thrombocytopenia	33 (1.2%)	22 (4.7%)	< 0.001	3.25	1.78–5.96	<0.001
	AMI	5 (0.2%)	12 (2.5%)	< 0.001	9.13	2.97–27.94	<0.001
	Pneumonia	21 (0.8%)	21 (4.4%)	< 0.001	3.56	1.76–7.17	< 0.001
	Urinary tract infection	77 (2.8%)	46 (9.7%)	< 0.001	2.05	1.21–3.47	0.008
	Postoperative delirium	2 (0.1%)	5 (1.1%)	< 0.001	10.17	1.78–57.97	0.009
	DVT	13 (0.5%)	10 (2.1%)	< 0.001	2.64	1.03–6.80	0.044
	Genitourinary disease	147 (5.4%)	69 (14.6%)	< 0.001	1.54	0.87–2.73	0.139
	Acute renal failure	74 (2.7%)	37 (7.8%)	< 0.001	1.15	0.58–2.31	0.688
	Gastrointestinal complication	53 (2.0%)	22 (4.7%)	< 0.001	1.70	0.95–3.03	0.072
	Convulsion	4 (0.1%)	2 (0.4%)	0.258	2.51	0.42–15.17	0.316
	PE	15 (0.6%)	6 (1.3%)	0.142	1.29	0.43–3.87	0.649
	Respiratory disease	19 (0.7%)	9 (1.9%)	0.020	1.20	0.47–3.06	0.706
	Septicemia	27 (1.0%)	19 (4.0%)	< 0.001	0.96	0.44–2.11	0.917
	GI bleeding	3 (0.1%)	1 (0.2%)	0.474	0.10	0.00-4.67	0.240
	Chest pain	14 (0.5%)	1 (0.2%)	0.597	0.21	0.02–1.91	0.164
	Urinary retention	76 (2.8%)	14 (3.0%)	0.846	0.95	0.52–1.75	0.875
	Respiratory failure	10 (0.4%)	8 (1.7%)	0.001	1.53	0.48–4.86	0.467
	Continuous trauma ventilation	8 (0.3%)	8 (1.7%)	< 0.001	2.11	0.62–7.16	0.230
	Arrhythmia	4 (0.1%)	2 (0.4%)	0.220	3.39	0.61–18.89	0.164
	Heart failure	14 (0.5%)	9 (1.9%)	0.030	1.74	0.64–4.73	0.277
	Cardiac arrest	5 (0.2%)	4 (0.8%)	0.042	2.08	0.36–12.22	0.416
	Postoperative shock	4 (0.1%)	5 (1.1%)	0.382	1.66	0.34–8.02	0.582
Surgical complications
	Wound infection	18 (0.7%)	16 (3.4%)	< 0.001	2.62	1.23–5.62	0.013
	Surgical incision dehiscence	8 (0.3%)	10 (2.1%)	< 0.001	2.52	0.83–7.69	0.103
	Urinary tract injury	4 (0.1%)	2 (0.4%)	0.220	3.19	0.56–18.33	0.193
	Lower limb peripheral nerve injuries	16 (0.6%)	6 (1.3%)	0.178	2.01	0.73–5.58	0.178

OR: Odds ratio, CI: Confidence interval, DVT: Deep vein thrombosis, PE: pulmonary embolism, AMI: Acute myocardial infarction, GI bleeding: Gastrointestinal bleeding

We also conducted region of hospital-stratified regression analyses to further identify the possible risk factors for blood transfusion. Notably, non-elective surgery emerged as a significant predictor of blood transfusion in most regions, except in the West hospital (Supplementary Fig. 3). Also, in the Northeast region, the initial logistic regression model, devoid of interaction terms, indicated that black patients were more likely to undergo blood transfusions compared to their white counterparts (P = 0.007). Also, in the northern region after considering the interaction between race and payment type there is no significant difference between the black and the white (P = 0.054).(Fig. 3).

Fig. 3.

Risk factors associated with blood transfusion after abdominal radical hysterectomy in Northeast

After PSM, the Black people is no longer a significant risk factor(P = 0.136). (Supplementary Table 5)

Discussion

In the United States alone, an estimated 13,960 individuals receive a cervical cancer diagnosis annually, with approximately 4,310 succumbing to the disease in 2023 [16]. Treatment for cervical cancer is primarily determined by the stage of the cancer and often involves a combination of chemoradiation and surgery, or surgery alone. And since 2018, several studies have suggested that minimally invasive surgery may be associated with lower DFS and OS rates, along with a higher risk of recurrence compared to the traditional open surgical approach [3, 17]. This has led to a widespread adoption of ARH in recent years. Consequently, this study was aimed at exploring the incidence and possible factors associated with ARH.

In this study, 473 patients (14.84%) received blood transfusions during perioperative period. Notably, blood transfusion rates exhibited a declining trend, decreasing from 18.83% in 2010 to 9.76% in 2019 (Fig. 2). The transfusion rate during ARH identified in this study was higher than that reported in another study, which documented a blood transfusion incidence of 10.7% [18]. This discrepancy is likely due to differences in the data collection period, as the surgical techniques and clinical guidelines may have evolved between these periods [18]. As expected, the blood transfusion rates in the ARH group substantially exceeded the rate in the minimally invasive surgery group, consistent with findings from other studies [15, 18, 19]. However, the mortality rate among patients who received transfusions during ARH was not significantly different from that reported laparoscopic surgery transfusion, suggesting that transfusions themselves may not be the primary driver of mortality differences between these approaches [15]. Additionally, a study published in 2021 found that the incidences of any single complication, intraoperative complications, and postoperative complications were higher in the LRH group compared to the ARH group [19].

Furthermore, limited research has investigated the predictors of blood transfusion for this patient population [20]. Our comprehensive analysis, spanning the years 2010 to 2019, examines both the economic and healthcare implications of perioperative blood transfusion in ARH. We employed logistic regression analysis to explore the association between hospital characteristics, demographic variables, and the likelihood of blood transfusion. As highlighted previously, our study identified a higher likelihood of blood transfusion among patients of Asian, Pacific Islander, or Black descent. This may be attributable to the documented high proportion of anemics within these populations [21]. Interestingly, an increased propensity for administering blood transfusions was observed in rural hospitals located in the Northeastern region of the United States. This phenomenon may be attributed to several factors, including disparities in available medical technology, variations in institutional protocols, and adherence to differing transfusion standards. Specifically, the criteria utilized to determine the necessity for transfusions in this region may play a crucial role in influencing transfusion practices [15, 22, 23].

Furthermore, patients with private insurance may exhibit a higher socioeconomic status and potentially healthier baseline profiles [24, 25]. This could translate to a lower prevalence of conditions like coagulopathy, chronic blood loss anemia, and nutritional deficiencies, consequently reducing the need for blood transfusions. Additionally, our findings suggest that elderly patients are more susceptible to requiring blood transfusion. This could be due to their reduced tolerance for blood loss and a higher burden of cardiovascular comorbidities [26].

It is now understood that several underlying health conditions significantly elevate a patient’s need for blood transfusions. These conditions, often characterized by diminished hemoglobin levels and increased bleeding risks, include coagulopathies, chronic and deficiency anemias, and fluid and electrolyte imbalances. Extensive data analysis supports this link, suggesting that blood loss, hemorrhages, shock, and gastrointestinal bleeding can all contribute to these imbalances and necessitate transfusions [27].

Our study further identified patients with peripheral vascular disorders, and pulmonary circulation disorders as at-risk groups. Chronic complications of diabetes can also increase transfusion risk. These conditions likely elevate transfusion needs due to a combination of factors, including changes in blood vessel permeability, iron or protein deficiency anemia, and nutrient depletion associated with chronic illness [28, 29].

Research indicates significant correlations between certain underlying health conditions and patients’ need for blood transfusions. These conditions are typically associated with reduced hemoglobin levels and increased bleeding tendencies, including coagulation disorders, chronic and deficiency anemias, and fluid and electrolyte imbalances. Extensive data analysis supports this association, suggesting that blood loss, hemorrhage, shock, and gastrointestinal bleeding may all be linked to these imbalanced states and transfusion requirements [28].

Our study also found that patients with peripheral vascular disease, and pulmonary circulation disorders may exhibit higher correlations with transfusion needs. Chronic complications of diabetes also demonstrated correlations with transfusion requirements. The association of these conditions with transfusion needs may involve multiple factors, including alterations in vascular permeability, iron or protein deficiency anemias, and nutritional depletion associated with chronic illnesses.

Our study observed that in certain regions, black patients were more likely than white patients to receive blood transfusions, but this difference was no longer significant after considering the interaction between race and payment type. This may reflect the complex relationship between race and socioeconomic status [30, 31]. On one hand, different payment types may represent varying economic levels and access to medical services among patients, potentially influencing physicians’ decisions regarding blood transfusion. On the other hand, racial disparities might be associated with unequal distribution of medical resources, differences in quality and access to healthcare services [32]. Research indicates that disparities in cervical cancer rates and outcomes are attributed to unequal access to healthcare services and the insufficient awareness, knowledge, and perceived susceptibility to HPV and cervical cancer among vulnerable women. It is crucial to advocate for preventive measures across the broader population, particularly for vulnerable groups, as such initiatives can not only reduce the incidence of complications and comorbidities but also lower the rates of cervical cancer itself.

This study has several limitations inherent to research using large administrative databases. The NIS primarily captures healthcare utilization data, lacking clinically relevant details such as cervical cancer stage, blood loss, transfusion amounts, anesthesia types, and perioperative hemoglobin levels [19]. Thus, a prospective cohort study is needed to further identify the potential relationship between the blood transfusion and other complications. Although this limits the depth of clinical analysis, it does not impede the identification of risk factors. A key strength of the NIS is its large sample size, which strengthens the findings and offers valuable insights for managing invasive cervical cancer patients undergoing ARH. Additionally, the NIS focuses on hospitalization data, omitting long-term post-discharge complications [12]. This limits our ability to track outcomes beyond the inpatient stay. Coding inaccuracies or omissions related to complications and transfusions may also affect the results, emphasizing the need for accurate coding practices. Finally, as with any observational study, unmeasured confounding factors may influence outcomes. However, the large sample size helps mitigate potential biases.

Conclusion

Our study underscores the increasing challenge of perioperative blood transfusion in radical hysterectomy, particularly with the rise in open surgeries. Blood transfusion was associated with significant postoperative complications, emphasizing the need for gynecologic oncologists to adopt strategies that minimize intraoperative blood loss. Key patient factors, such as race, insurance type, and pre-existing health conditions, should guide preoperative optimization to improve outcomes and reduce unnecessary complications.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Chuan Chen: Formal analysis, Writing – original draft, Writing – review & editing; BenLi Zhu: Data curation, Writing – original draft, Writing – review & editing; YouFeng Wang: Data curation, Writing – original draft; YangYang Zhao: Figurse and tables design; Gang Chen: Conceptualization, Project administration; Ying Peng: Conceptualization, Methodology; Ying2 Peng: Data curation, Investigation, Resources, Software; XinYu Wang: Conceptualization, Resources, Software; Hao Xie: Methodology, Supervision, Validation; Ying Zhou: Conceptualization, Supervision, Validation, Visualization; Juan Lin: Methodology, Supervision, Validation, Visualization, Revision.

Funding

This work was supported by Guangzhou Science and Technology Plan Project, Number 2023A04J0624.

Data availability

The datasets are available at https://www.ahrq.gov/data/hcup/index.html.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.