Clinical value of c-reactive protein level in predicting the development of postpartum endometritis

Full Text

Introduction

Purulent-inflammatory diseases are urgent problems in modern obstetrics [1–7]. The introduction of antibiotics into obstetric practice more than half a century ago rapidly decreased the incidence of postpartum infectious diseases. However, in the last decade, the incidence of postpartum infections increased worldwide [2–4]. The incidence of postpartum infectious diseases varies from 2% to 10% because of the lack of uniform diagnostic criteria [1, 2, 6]. Infectious complications develop after cesarean section and occur in 20% of patients [1, 3, 4, 6, 8]. In the Russian Federation, septic complications in the postpartum period rank third among the causes of maternal mortality along with obstetric hemorrhage and preeclampsia [9]. Many factors, such as severe extragenital pathology of pregnancy, assisted pregnancy, surgical correction of miscarriage, and changes in the nature of microflora, contribute to the growth of purulent–septic pathology [1, 7, 10].

The widespread and misuse of broad-spectrum antibiotics have caused the emergence of bacterial strains with multiple resistance to antibacterial drugs. As a result of natural selection, weak microorganisms with low resistance to unfavorable conditions disappear, and antibiotic-resistant species and strains accumulate in clinics [6, 8]. Large obstetric hospitals play a negative role in the prevention of postpartum infectious diseases because the high concentration of pregnant women, postpartum women, and newborns, which are susceptible to infection given their physiological characteristics, increases the risk of purulent–septic complications [5, 11–13]. Two factors increasing infectious complications in obstetric practice are the widespread use of invasive diagnostic methods (fetoscopy, amniocentesis, cordocentesis, direct fetal electrocardiography, and intrauterine tocography) and the introduction of surgical aids in pregnant women (surgical correction of istmico–cervical insufficiency in miscarriage) [3, 4, 8, 11, 12]. Early diagnosis of infectious complications is also difficult because of the wide spread of unclear forms and late clinical manifestations [6, 14].

Measurement of the protein content in the acute phase of inflammation is reasonable to assess the risk of development and the severity of the course of the pathological process in the postpartum period [4, 15]. Commonly used laboratory markers of inflammation include C-reactive protein (CRP), procalcitonin, and presepsin [3, 4, 13, 16, 17]. Despite its high sensitivity and specificity, procalcitonin is used in the Clinical Guidelines for the Initial Treatment of Sepsis and Septic Shock in Obstetrics (2018) as a diagnostic criterion for an already developed systemic inflammatory response syndrome to predict its severity and as a criterion in making a decision on hysterectomy and antibiotic therapy withdrawal (level B-IIa) [3–4, 7, 16]. In addition, procalcitonin and CRP, which have the same levels of evidence significance and recommendation strength, are used as criteria for assessing the quality of care provided to a patient with sepsis and septic shock [7] (Table 1).

 

Table 1 / Таблица 1

Criteria (indicators) for assessing the quality of medical care provided to a patient with sepsis and septic shock

Критерии (индикаторы) оценки качества медицинской помощи, оказанной пациентке с сепсисом и септическим шоком

No.	Quality criteria	Level of evidence	Strength of recommendation
6	C-reactive protein and/or procalcitonin level measured per hour (yes/no)	A	I

 

Presepsin is used only as a predictor of the outcome and severity of inflammatory response in the diagnosis of sepsis; it is not included in the clinical protocol for the initial therapy of sepsis and septic shock and is currently being finalized [4, 7, 10, 11, 13]. At the same time, the level of evidence and the grade of recommendations have not been established for presepsin. CRP is referred to as the “main” protein of the acute phase of inflammation in humans because its level increases rapidly (in the first 6–8 h) and significantly (20–100 times) [3]. CRP has been known for over 100 years as the gold standard marker of inflammation, and its determination changed from semi-quantitative to high-precision methods.

A consensus about the timing of postpartum CRP determination as a pro-inflammatory marker for the early diagnosis of infectious complications remains lacking to date [2]. In healthy puerperas, the CRP level moderately increases by day 3 of the postpartum period. In the puerperas with endometritis, the CRP level increases in the preclinical stage of the onset of the disease symptoms. This feature is prognostically considered an unfavorable and early sign of the disease, correlating directly with the severity of its course. However, the timing of CRP determination and the rate of its content in the blood serum of puerperas are unknown [1]. A continuous increase in CRP levels in patients with suspected sepsis within 120 h is associated with an increased risk of poor outcome. In addition, the dynamics of an increase in the CRP level in sepsis correlates with the severity of the condition and the dysfunction of systems and organs [3]. CRP is an important predictor of the development of sepsis, but further study of this issue is required because of the ambiguity of the literature data.

The work aimed to assess the clinical informational value of determining the CRP level in predicting the development of postpartum endometritis.

Materials and methods

The study included 135 with a high risk of purulent–septic complications. The puerperas were retrospectively distributed into two groups. The main group included puerperas with postpartum endometritis (n = 72), and the comparison group included puerperas with the physiological course of the postpartum period (n = 63).

Postpartum endometritis was diagnosed in accordance with the International Statistical Classification of Diseases and Problems Related to Health, 10^th revision and was presented in accordance with the diagnostic criteria for postpartum endometritis, formulated in the clinical guidelines “Purulent-inflammatory diseases and sepsis in obstetrics” of the Ministry of Health of the Russian Federation (2015). For diagnosis of postpartum endometritis, two of the following criteria are sufficient:

1. an increase in body temperature (38°C and above);
2. soreness of the uterus on palpation;
3. discharge from the genital tract with an unpleasant odor;
4. moderate bleeding from the genital tract.

Criteria for inclusion in the main group were as follows:

5. risk factors of purulent–septic complications;
6. informed voluntary consent to participate in the study.

Exclusion criteria were as follows:

7. physiological course of the postpartum period;
8. other purulent–septic complications of the postpartum period without endometritis (infection of the surgical site, lactational mastitis).

Withdrawal criteria included the following:

9. decompensated extragenital pathology;
10. the use of glucocorticosteroids during pregnancy and in the postpartum period;
11. refusal to take part in the study.

All puerperas were monitored daily until discharge from the hospital. The general condition was assessed, complaints were taken into account, and signs of systemic inflammatory reaction syndrome were identified and recorded in the delivery history. On day 1 after childbirth, the anamnesis of the parturient women was assessed, the course of pregnancy and delivery was analyzed, and informed consent was obtained for participation in the study. On days 1 and 3 of the postpartum period, the CRP level in the blood serum was determined in all female patients through immunoturbidimetry (automatic biochemical analyzer Vitros 350, France, Johnson & Johnson reagents, Great Britain).

The average age of postpartum women was 31.3 years (confidence interval (CI) 30.0–32.1) in the main group and 31.2 years (CI 30.3–32.1) in the comparison group (p = 0.96). The main group comprised 46% of primigravida patients (33 puerperas), whereas the comparison group comprised 36% of primigravida patients (46 puerperas). The proportion of multigravida patients was 54% (39 puerperas) in the main group and 63% (80 puerperas) in the comparison group (p = 0.2). In the groups under study, pregnancy ended in term labor in 76% of puerperas in the main group and in 66% of puerperas in the comparison group. On the basis of the somatic, obstetric, and gynecological anamnesis and the course of this pregnancy and childbirth, the groups of puerperas did not differ statistically significantly (Table 2). All postpartum women enrolled in the study had compromised somatic and obstetric–gynecological anamnesis.

 

Table 2 / Таблица 2

Frequency of risk factors for purulent-septic complications in the study groups

Частота факторов риска развития гнойно-септических осложнений в исследуемых группах родильниц

Risk factor	Main group (n = 72)	Comparison group (n = 63)
Extragenital pathology	62 (86.1%)	63 (100%)
Gynecological pathology
Chronic urogenital infections	35 (48.6%)	33 (48%)
Inflammatory diseases of the pelvic organs	11 (15.3%)	6 (9%)
Abortions	27 (37.5%)	29 (46%)
Infertility	16 (22%)	14 (22%)
Complications of pregnancy
Threatening miscarriage	29 (40.3%)	19 (30.1%)
Cervical insufficiency	12 (16.7%)	7 (11.1%)
Gestational diabetes mellitus	16 (22.2%)	13 (20.5%)
Gestational pyelonephritis	2 (2.7%)	3 (4.8%)
Asymptomatic bacteriuria	7 (9.7%)	3 (4.8%)
Complications of labor
Premature delivery	17 (23.6%)	16 (25.3%)
Premature rupture of membranes	21 (29.2%)	21 (33.3%)
Prolonged period without amniotic fluid	11 (16%)	15 (23%)
Chorioamnionitis	10 (13.9%)	11 (17.4%)
Cesarean section	48 (65%)	17 (27%)*
Technical aspects when performing a cesarean section	25 (35%)	2 (6%)*
Surgical interventions during childbirth	16 (22%)	34 (54%)*

Note: * statistically significant differences.

 

Parturient women in the main group delivered by cesarean section 2.4 times more often than the healthy parturient women (48 parturient women in the main group (65%; 95% CI 55–75) and 34 parturient women in the comparison group) (27%; 95% CI 20–35) (RR = 2.4, OR = 5.0, p = 3 · 10⁻⁸). The frequency of cesarean sections performed in a planned manner showed no significant difference between the groups (11 puerperas in the main group (15%; 95% CI 7–28) and 11 puerperas in the comparison group) (9%; 95% CI 4–16). The proportion of cesarean sections performed on an emergency basis was significantly higher in the main group (37 puerperas) (51%; 95% CI 37–66) than in the comparison group. In the comparison group, a cesarean section was performed in 23 puerperas (18%; 95% CI 11–28) (p = 10⁻⁷). Delivery by cesarean section increased the risk of postpartum endometritis by five times. The technical aspects of the cesarean section (duration of the surgery of more than an hour, additional surgical hemostasis, ligation of the ascending uterine arteries, expansion of the surgery amount) played an equally important role in the development of postpartum endometritis. Technical difficulties that arose during the cesarean section were 5.8 times more common in the postpartum women in the main group (25 postpartum women in the main group (35%; 95% CI 25–46) and 8 patients in the comparison group) (6%; 95% CI 3–12) (RR = 5.8, OR = 7.5, p = 3 ∙ 10⁻⁷), which increased the risk of postpartum endometritis by 7.5 times. No significant difference in the frequency of chorioamnionitis and in the anhydrous period was found between the groups of postpartum women.

In the main group, clinical manifestations of endometritis were recorded on day 4.8 (95% CI 4.5–5.1) after spontaneous delivery, on day 3.4 (95% CI 2.3–4.0) after cesarean section (p = 10 ∙ 10⁻⁷), and on day 4.4 (95% CI 3.4–5.4) in the presence of chorioamnionitis in labor (p = 0.28).

A symptom of a temperature reaction (more than 38°C and higher) was noted in 54 puerperas (75%), pain on palpation of the uterus was detected in 54 (75%), and the pathological nature of lochia was noted in 60 patients (83%). All three diagnostic criteria for postpartum endometritis were recorded in 39 puerperas (54%), and the diagnosis was made in accordance with two criteria in 33 puerperas (46%).

In 41 puerperas of the main group (60%), isolated endometritis was established. In 31 puerperas (40%), endometritis was registered in combination with other purulent–septic complications. In the presence of postpartum endometritis, an infection of the surgical site (infiltrates of the sutures of the perineum, anterior abdominal wall after cesarean section, seroma of postoperative sutures) was recorded in nine puerperas, eight puerperas had pelvic hematomas, and 11 puerperas had subgaleal hematomas. In addition, in seven female patients (23%), postpartum endometritis was complicated by the development of severe purulent–septic diseases, namely, sepsis in three puerperas, pelvioperitonitis in two patients, and parametritis in two patients.

The clinical informational value of the method was assessed by calculating the standardized effect size (SES) index and by performing ROC analysis. The main indicators of the quality and informational value of the diagnostic test were sensitivity, specificity, likelihood ratio, and predictability. For all indicators, the 95% CI was calculated.

Results

On day 1 of the postpartum period, the diagnostic threshold value of CRP in the puerperas was 69 mg/mL. The sensitivity and specificity of the method on day 1 of the postpartum period were low, 62% (95% CI 0.5–0.74) and 65% (95% CI 0.51–0.76), respectively. The predictive value for CRP levels above 69 mg/mL was also low and amounted to 67% (95% CI 54–77). As in the puerperas on day 1 of the postpartum period with a CRP level above 69 mg/mL, the probability of endometritis was 67%. In addition, the likelihood ratio for the diagnostic CRP level was 1.76. That is, the chances of endometritis development in the puerperas on day 1 of the postpartum period with a CRP level above 69 mg/mL increased by 1.76 times (95% CI 1.2–2.58). Furthermore, the likelihood ratios were extremely low on the basis of the scale of verbal interpretation.

Thus, the determination of CRP has no diagnostic value on day 1 of the postpartum period.

On day 3 of the postpartum period, the diagnostic threshold value of CRP was 60 mg/mL. The method had a moderate sensitivity of 79% (95% CI 68–86) and a high specificity of 93% (95% CI 85–98). The predictive value for CRP levels above 60 mg/mL was 93% (95% CI 84–96). In the puerperas on day 3 of the postpartum period with CRP levels above 60 mg/mL, the probability of endometritis was 93%. The likelihood ratio for the diagnostic CRP level was 12.5, that is, the probability of endometritis in the puerperas on day 3 of the postpartum period with a CRP level above 60 mg/mL increased by 12.5 times (95% CI 5–30). Furthermore, the likelihood ratios were considered moderate and clinically significant on the basis of the criteria of the verbal interpretation scale (Table 3).

 

Table 3 / Таблица 3

Assessment of diagnostic informational content of C-reactive protein level in dynamics

Оценка диагностической информативности уровня С-реактивного белка в динамике

Indicator	Day 1 of the postpartum period	Day 3 of the postpartum period
Diagnostic threshold level, mg/ml	69	60
Sensitivity, %	50 62 74	68 79 86
Specificity, %	51 65 76	85 93 98
Predictability, %	54 67 77	84 93 96
Likelihood ratio at the level of C-reactive protein higher than diagnostic, LR [+]*	1.20 1.76 2.58	5 12.5 30

Note: * The likelihood ratio for a positive diagnostic test result (LR [+]) indicates how many times the probability (chances) of developing the disease is greater than the probability of its absence. 95% CI is indicated as subscripts.

 

Thus, the determination of CRP has a clinical information value and a high predictive value on day 3 of the postpartum period.

The CRP level on day 1 of the postpartum period showed no significant differences between the groups. On day 3 of the postpartum period, the CRP level was significantly higher in the puerperas in the main group than in those in the comparison group. In the patients of the main group, the CRP level on day 3 after delivery was 148 mg/mL (95% CI 126–171), whereas that in the comparison group was 43 mg/mL (95% CI 38–49) (SES = 1.4; p = 6 ∙ 10⁻¹⁴) (Table 4).

 

Table 4 / Таблица 4

C-reactive protein levels in the study groups on day 3 of the postpartum period

Содержание С-реактивного белка у родильниц исследуемых групп на третьи сутки послеродового периода

Indicator	Main group of puerperas (n = 72)	Comparison group (n = 63)	SES	p
C-reactive protein	126 148 171	38 43 49	1.0 1.4 1.8	6 ∙ 10–14

Note. SES — standardized effect size. 95% CI is indicated as subscripts.

 

In the patients of the main group, the CRP level varied depending on severity of postpartum endometritis. Thus, on day 1 of the postpartum period, the CRP levels were 110 mg/mL (95% CI 86–110) in the subgroup of postpartum women with isolated endometritis (n = 41), 100 mg/mL (95% CI 67–134) in the subgroup of postpartum women with endometritis in combination with another purulent–septic complication (n = 27), and 240 mg/mL (95% CI 112–350) in the subgroup of postpartum women with endometritis in combination with two or more purulent–septic complications, including severe forms (sepsis, parametritis, pelvioperitonitis), with the maximum value reaching 350 mg/mL. The CRP level in the subgroup of patients with severe purulent–septic diseases on day 1 of the postpartum period was significantly higher than that in the two other subgroups (p = 10⁻⁴, p = 0.003, SES = 3.02). On day 3 of the postpartum period, the CRP level was 146 mg/mL (95% CI 119–174) in the subgroup of puerperas with isolated endometritis, 138 mg/mL (95% CI 100–177) in the subgroup of puerperas with postpartum endometritis in combination with another purulent–septic complication, and 237 mg/mL (95% CI 173–280) in the subgroup of puerperas with endometritis in combination with two or more purulent–septic complications, including those with severe diseases, with the maximum value reaching 320 mg/mL. At the same time, the CRP level did not statistically significantly differ in the subgroups of patients on day 3 of the postpartum period (p = 0.95; p = 0.08; p = 0.05). As a result, the development of severe purulent–septic complications after childbirth was not predicted.

One of the most significant risk factors of postpartum endometritis is surgical interventions during childbirth, including a cesarean section. Cesarean section does not significantly affect the CRP level. Thus, on day 1 of the postpartum period, the CRP level was 107 mg/mL in the puerperas of the main group who underwent cesarean section and 75 mg/mL in the women after spontaneous delivery (p = 0.02). On day 3 of the postpartum period, the CRP level was 103 mg/mL in the patients of the main group who underwent cesarean section and 97 mg/mL in the patients after spontaneous delivery (p = 0.75).

To confirm the practical significance and clinical information value of determining the CRP level on day 3 of the postpartum period, we calculated the indicator of SES: SES = 1.4 (p = 6 ∙ 10⁻¹⁴) (Fig. 1). Considering that the SES value exceeds the clinically significant threshold value of 1, we can interpret the determination of the CRP level as a method characterized by high clinical significance and efficacy.

 

Fig. 1. Graphical analysis of CRP levels in the study groups on day 3 of the postpartum period. The 95% CI is indicated as subscripts

Рис. 1. Графический анализ уровня С-реактивного белка в группах родильниц на третьи сутки послеродового периода. 95 % ДИ указан в виде подстрочных индексов

 

This conclusion was also confirmed by the ROC analysis results (Fig. 2). The numerical value of clinical significance (AUC indicator) was 0.88 (CI 0.81–0.93), suggesting that the definition of CRP is clinically informative.

 

Fig. 2. ROC analysis of the evaluation of clinical informativity of C-reactive protein level. The 95% CI is indicated as subscripts

Рис. 2. ROC-анализ оценки клинической информативности уровня С-реактивного белка. 95 % ДИ указан в виде подстрочных индексов

 

Discussion

The relevance of postpartum purulent–septic diseases, including postpartum endometritis, is beyond dispute. Thus, this study aimed to search for new risk factors of postpartum purulent–septic complications, as well as their combinations. Postpartum endometritis develops in case of a compromised somatic and obstetric–gynecological history, which causes an early onset of clinical signs, namely, on 4.8 days after spontaneous delivery and on 3.4 days after a cesarean section [4, 7].

Our results showed cesarean section as the most significant risk factor of purulent–septic complications after childbirth and some technical aspects of its implementation, which is consistent with previous studies [6, 7, 10]. Cesarean section was performed significantly more frequently to patients of the main group (mostly on an emergency basis) than to patients of the comparison group. Cesarean section performed on an emergency basis increases the risk of postpartum endometritis by five times. Technical difficulties, such as the duration of the surgical intervention for more than an hour, additional surgical hemostasis, ligation of the ascending uterine arteries, and expansion of the extent of the surgery, increase the risk of postpartum endometritis by 7.5 times.

Our study revealed the threshold diagnostic values of CRP on days 1 and 3 of the postpartum period (69 and 60 mg/mL, respectively). With this regard, the CRP level below the specified diagnostic threshold value is the norm in puerperas with a high risk of purulent–septic complications. The available literature does not describe the CRP normal values for the postpartum period. On day 1 of the postpartum period, the determination of the CRP level does not have sufficient diagnostic value because the sensitivity, specificity, and predictive value of the method are low, as confirmed by the data of other authors [1, 3]. In addition, the absence of statistically significant differences in CRP levels in patients in the study groups on day 1 of the postpartum period is associated with CRP nonspecificity. Its increased level in postpartum women of both groups represents an adaptive response of the body to birth stress, surgery, etc. Therefore, it cannot be used in the early diagnosis of postpartum endometritis as a predictor. Nevertheless, despite the absence of a significant difference in CRP levels in the studied groups of puerperas on day 1 of the postpartum period and the lack of the necessary diagnostic value of this indicator, our data are consistent with the data of other authors that CRP levels higher than 200 mg/mL (200–350 mg/mL) are associated with the development of severe forms of purulent–septic complications. The CRP level of 200–300 mg/mL is characteristic of severe generalized infections and sepsis [3]. On day 3 of the postpartum period, the determination of CRP has moderate sensitivity (79%), high specificity (93%), and predictability (93%).

Numerous studies reported that the sensitivity of the procalcitonin test is 80%–85%, and that of presepsin is 87%–91%. The present results showed that the specificity of CRP on day 3 after delivery was comparable with those of procalcitonin (95%) and presepsin (98%), but these inflammatory laboratory markers cannot be used for routine diagnosis because of their limited availability and high cost [3, 16]. Previous studies have described a low sensitivity of CRP in postpartum women (66%) and a low diagnostic value of CRP in the early diagnosis of postpartum endometritis, especially after cesarean section [1, 6]. The authors argued that CRP cannot be used for the early diagnosis of purulent–septic complications and for differential diagnosis because of its nonspecificity and rapid increase. On the basis of many years of experience in the early diagnosis and treatment of postpartum endometritis in puerperas with a high risk of developing infectious complications in a large perinatal center, delivery by cesarean section and surgical interventions in childbirth (mainly manual examination of the uterine cavity) do not affect the level of CRP. In addition, statistical analysis revealed that CRP is informative to use as a predictor of postpartum endometritis in puerperas with a high risk of purulent–septic complications on day 3 of the postpartum period.

Conclusion

Determining CRP level on day 3 of the postpartum period is clinically informative. An increase in CRP levels of more than 60 mg/mL is a predictor of the development of postpartum endometritis with a sensitivity of 79% and a probability of 93%.

Additional information

Conflict of interest. The authors declare no obvious or potential conflicts of interest related to the publication of this article.

About the authors

Tatyana V. Batrakova

V.A. Almazov National Medical Research Center

Author for correspondence.
Email: tatyana_shuranov@mail.ru
ORCID iD: 0000-0002-5710-9416
SPIN-code: 3042-2856

MD, Post-Graduate Student. The Department of Obstetrics and Gynecology, the Faculty of Medicine, the Institute of Medical Education

Russian Federation, Saint Petersburg

Irina E. Zazerskaya

V.A. Almazov National Medical Research Center

Email: zazera@almazovcentre.com
ORCID iD: 0000-0003-4431-3917
SPIN-code: 5683-6741

MD, PhD, DSci (Medicine), the Head of the Department of Obstetrics and Gynecology. The Faculty of Medicine, the Institute of Medical Education

Russian Federation, Saint Petersburg

Tatyana V. Vavilova

V.A. Almazov National Medical Research Center

Email: Vavilova_TV@almazovcentre.ru
SPIN-code: 9003-6455

MD, PhD, DSci (Medicine), Professor, the Head of the Department of Laboratory Medicine and Genetics. The Faculty of Medicine, the Institute of Medical Education

Russian Federation, Saint Petersburg

Vitaly N. Kustarov

V.A. Almazov National Medical Research Center

Email: zazera@almazovcentre.com

MD, PhD, DSci (Medicine), Professor. The Department of Obstetrics and Gynecology, the Faculty of Medicine, the Institute of Medical Education

Russian Federation, Saint Petersburg

References

Supplementary files