The labor pain experience is a complex interaction between endogenous nociceptive receptor-induced pain due to the process of birth and the interpretation and expression of this pain stimulus. Labor pain is affected by many psychological, obstetric, and social factors beyond just nociceptive receptor-induced pain.1,2 The labor pain experience is particularly challenging to define because of the many unknown maternal and fetal factors (such as duration of labor, obstetric outcome, and neonatal well-being) that evolve during the labor and delivery process. A better understanding of prelabor factors that influence the labor pain experience might help improve intrapartum pain management by tailoring treatment to the woman’s analgesic and psychosocial requirements.
A number of psychological factors have been shown to impact postoperative pain and analgesic use.3,4 Studies conducted in a surgical setting have found that pain catastrophizing, anxiety, and fear of pain are significantly correlated with postoperative and experimental pain models.3–10 In addition, personality traits, especially neuroticism, have been found to influence reporting of pain.11,12 The impact of psychological factors on labor pain and analgesic use is not completely understood. Anxiety has been shown to negatively impact pain experienced during labor.13,14 Previous studies have found that pain catastrophizing correlates with postcesarean delivery pain,15 as well as postpartum depression and functional recovery after delivery.16 The role of psychological factors, such as fear of pain, pain catastrophizing and personality traits, and their combined or interactive effects on the labor pain experience, and maternal satisfaction with labor analgesia are unknown.
The primary aims of this study were to determine whether key psychological characteristics (fear of pain, anxiety, pain catastrophizing, and personality traits) measured using validated questionnaires and 3-scaled ratings of anxiety, confidence, and analgesic expectations determined before the onset of labor predict labor pain scores, epidural local anesthetic use, and maternal satisfaction with labor analgesia. The secondary objectives were to investigate the relationship among these important psychological factors. We hypothesized that these psychological tests, if administered before the onset of labor, would predict pain, analgesic use, and analgesic satisfaction during labor.
Approval for this prospective observational case-controlled study was obtained from the Stanford University IRB. Fifty healthy (ASA physical status 1 or 2) nulliparous women aged 18 to 40 years, with a singleton term or post-term (37–42 weeks’ gestation) pregnancy, admitted for scheduled induction of labor and planning to request labor epidural analgesia were enrolled in the study. Patients were approached, and written informed consent was obtained before the induction and onset of labor. The study was conducted at Lucile Packard Children’s Hospital, California. Exclusion criteria included significant medical or obstetric disease, trial of labor after cesarean delivery, morbid obesity, inability to understand or read English (validated psychological tests were in English), history of chronic opioid or antidepressant use, recent (<48 hours) intake of any analgesic medications, opioid or local anesthetic intolerance or allergy, and any contraindication for epidural analgesia. A priori we selected to only analyze the data of patients undergoing successful vaginal delivery and exclude those who progressed to intrapartum cesarean delivery.
Prelabor Assessment of Psychological Predictors for Labor Pain
Four validated (in a nonlabor setting) psychological questionnaires and 3 simple rating scale questions were administered to study subjects on admission to the labor and delivery ward and before commencing induction of labor. All the questionnaires were self-administered by the participants (in the order outlined below). The psychological questionnaires were as follows:
- An Anxiety Sensitivity Index (ASI): A 16-question written questionnaire that assesses the degree of anxiety associated with potentially unpleasant events and determines possible negative consequences related to the experience of anxiety.17 Each question is scored as very little, a little, some, much, and very much (0–4 points for each option) and the responses summated to a total score.
- A Fear of Pain Score (FPQIII): A 30-question written questionnaire that assesses fear generated from physical insults or injuries.18,19 Each question assesses the amount of fear (1 = not at all, 2 = a little, 3 = a fair amount, 4 = very much, and 5 = extreme), and the responses are summated to a total score.
- A Pain Catastrophizing Scale (PCS): A 13-question written questionnaire with 3 components (rumination, magnification, and helplessness). Each question’s options include not at all, to a slight degree, to a moderate degree, to a great degree, and all the time. All components are scored separately, and the total PCS score is calculated as the sum of the 3 component scores.20
- An Eysenck Personality Questionnaire–Short Scale (EPQR-S): A 48-question written questionnaire with 12 questions in each of 4 personality categories (psychoticism, extroversion, neuroticism, and lying).21,22 Each question answer (0 = no, 1 = yes) is summated to produce scores in each category. The higher the score is in each of the categories, the greater the tendency is toward that personality trait.
In addition to the validated questionnaires outlined above, 3 simple-scaled rating questions with the potential to predict the labor pain experience were also obtained on a separate written questionnaire: (1) How anxious are you about experiencing labor and labor pain? (0 = no anxiety, 100 = extreme anxiety); (2) How confident are you that you will be able to cope with labor pain? (0 to 100% confident); (3) What is your expected analgesic requirement? (0 = no analgesia, 10 = highest possible amount). Demographic (age, height, weight, race/ethnicity) and obstetric data (parity, induction indication, any expected fetal problems) were recorded. We determined whether women had previously undergone surgery and asked them to rate their most physically painful life experience (numerical verbal pain scale 0–10, 0 = no pain, 10 = worse pain imaginable).
Induction of labor was not standardized for the study protocol and followed hospital policy and obstetrician preference. Induction of labor at our institution typically includes misoprostol 50 μg orally, then 100 μg orally every 4 hours or dinoprostone 10 mg vaginally for cervical ripening, followed by increasing doses of oxytocin (starting at 1 mU/min and increasing by 2 mU/min every 30 minutes to a maximum of 30 mU/min). Pain scores during labor were assessed hourly from the onset of painful contractions until delivery, using a numerical verbal pain scale 0 to 10 (0 = no pain, 10 = worst pain imaginable). Pain scores were also assessed at maternal request for epidural labor analgesia. Epidural analgesia was administered as per standard care at our institution as follows: Patients were placed in the sitting position. A loss-of-resistance to saline technique was used to identify the epidural space at the L2-3 or L3-4 interspace, and a 19-G, single-orifice, spring-wound epidural catheter was inserted 5 cm into the space. Initial analgesia was provided with 15 mL of 0.125% bupivacaine plus sufentanil 10 μg injected in 5-mL increments administered at 3-minute intervals. Labor analgesia was maintained with patient-controlled epidural analgesia (PCEA) consisting of 0.0625% bupivacaine plus sufentanil 0.4 μg/mL at the following settings: 12 mL/h background infusion, 12-mL bolus, and 15-minute lockout interval. At the discretion of the anesthesiologist caring for the patient, 5 to 10 mL of 0.125% to 0.25% bupivacaine was administered for breakthrough pain and repeated as clinically indicated. These boluses were recorded on the patient’s record and were included in the data collection. After delivery, data from the PCEA pump were downloaded using Curlin CMS Gold software (Curlin Medical Clinical Management Systems, Huntington Beach, CA). The total amount of local anesthetic/opioid epidural solution used was determined by adding the PCEA consumption to the physician rescue boluses and recorded in milligrams per hour of analgesia. The epidural infusion stop time was defined as time of placenta delivery.
Response Measures and Study Outcomes
Key outcome measures chosen to reflect the labor pain experience were as follows: Time from onset of labor to epidural analgesia request (minutes), pain at epidural analgesia request, area under the pain × time curve (AUC) during labor, epidural local anesthetic consumption (mg/h), and overall maternal satisfaction with labor analgesia (0–100, 0 = totally unsatisfied, 100 = totally satisfied, recorded immediately after delivery). The onset of labor was defined as the onset of painful contractions as reported by the study subject. Study investigators recording response data were blinded to the preinduction questionnaire scores.
Demographic and outcome data are summarized as mean ± SD, median (interquartile range), and number (percentage) as appropriate. To check which pairs showed significant bivariate correlations, the Spearman ρ statistic was initially performed for each of the predictor and response pairs, and the method proposed by Benjamini and Hochberg23 was used to adjust for false discovery rate when multiple tests were performed simultaneously. Adjusted P < 0.05 was considered statistically significant. We applied a widely used convention to describe the magnitude of observed correlation r (both positive or negative): weak or no correlation if r < 0.4, moderate correlation if r = 0.4 to 0.7 and strong correlation if r > 0.7. The 95% confidence intervals (CIs) for the Spearman correlations were evaluated using the Fisher z-transformation.24 To control the false discovery rate, adjusted t statistics were derived from the adjusted P values, and adjusted correlation values were deduced accordingly. The primary predictors were ASI, FPQIII, overall PCS, EPQR-S personalities (psychoticism, extroversion, neuroticism, and lying), and the 3 simple rating questions (anxiety, confidence, and analgesic expectations). The key response variables were time from onset of labor to epidural analgesia request; pain at epidural analgesia request; labor pain × time AUC; hourly epidural local anesthetic consumption and satisfaction with labor analgesia. In addition, we performed bivariate analysis to determine the correlation between ASI and anxiety (0–100 scale) described above.
To achieve as close to linearity as possible between a predictor and a response variable, transformation of the predictor variables was performed, and the transformation that led to the largest absolute correlation between the transformed predictor and the response variable was used for multivariate linear regression analysis. The candidates for transforms were of the Box-Cox type: x 3, x 2.9, ..., x 0.1, log(1+x), 1/(1+x)0.1, ..., 1/(1+x)3. Because the predictors may have zero values, we uniformly offset the values by one before applying log-transformation or power transformation with negative power.
A forward–backward model selection method was selected for the multivariate linear regression modeling. To minimize overfitting, a selection method based on Akaike information criterion (AIC)25 was performed for each response variable to determine the best model (a balance between model fitting and model complexity) and statistical significance at P ≤ 0.01. Specifically, this model selection starts with a null model with no variables in the model. After each iteration, the model considers whether adding any variable that was not in the model (if any) or eliminating any variable currently included in the model (if any) would lead to a smaller AIC score. The process continued until no change further reduced the AIC score, and the final model was selected. The particular response and statistical significance were assessed for this final model only. Using this model selection method, multicollinearity among predictors in the model was minimal. The addition of another variable that correlated with the existing variable was unlikely to have a large effect in improving the model fitting and therefore would not be included by the model selection method. The statistical assumption of the multivariate linear regression model was examined graphically, as well as using the Shapiro-Wilk test for normality,26 to make sure the residuals were normally distributed and had roughly constant mean (=0) and variance across the entire range of the fitted values. Observations with outlying response variable values were excluded during the multivariate linear regression analysis to avoid spurious findings. Outlying values were determined by the following definition: below Q1 − 1.5 × interval quartile range or above Q3 + 1.5 × interval quartile range (where Q1 and Q3 are the first and third quartiles, respectively). The number of outliers removed for the response variables time to epidural analgesia request, pain at epidural analgesia request, labor pain AUC, hourly epidural local anesthetic consumption, and maternal satisfaction with labor were 2, 0, 2, 4, and 3, respectively.
A resampling analysis was implemented to investigate the robustness of the model fitting. For each iteration, three-fourths of the data (before removing outliers) were randomly sampled (without replacement). The outlier detection and model selection procedures were performed as done for the full data set. One thousand iterations were performed, and the number of times out of the 1000 iterations that the predicting variable was included in the final model for the selected response measurement using the forward–backward method was recorded (i.e., the “importance of count” of a predictor for that response measurement). The average P value of the model fitting, the average adjusted R 2 of the final model, the average P value of the Shapiro test (to test that residuals followed a normal distribution), and the average number of outliers excluded in the 1000 iterations were recorded.
To compare the relative contribution to labor pain AUC among ASI, FPQIII, and PCS, we used an F test on the nested multivariate linear regression model to assess the effect of adding ASI, PCS, or FPQIII to the model with all other 7 predictor variables (4 EPQR-S personalities of psychoticism, extroversion, neuroticism, and lying and 3 simple rating questions of anxiety, confidence, and analgesic expectations) present. The reduced model was the model with the other 7 variables, and the full model was the model with 1 of the variables of interest (ASI, PCS, or FPQIII) added into the model. An F test was used to compare the full model with the reduced model to evaluate the significance of the contribution of the added variable. We applied the same transformation and model selection method to select the relevant variables among the other 7 predictors to include in the final model; a P value of ≤0.05 was required for any of these 7 variables to be included in the model. The partial correlation of ASI-AUC and PCS-AUC (or FPQIII-AUC) with the presence of the other 7 predictors was evaluated using the “ppcor” package (http://cran.r-project.org/web/packages/ppcor/index.html).
A permutation test was implemented to assess the statistical significance among the observed difference in correlations (ASI-AUC, PCS-AUC, and FPQIII-AUC). We used both bivariate Spearman correlation (that did not consider other variables) and the partial correlation (that characterized the relationship after the effects from other variables were considered). In the permutation test, the values for ASI, PCS, and FPQIII together were randomly permuted, while values for AUC and the other 7 predictors remained unchanged. By doing so, the correlation among the variables of interest (ASI, PCS, and FPQIII) and the response variable (AUC) and the other predictors remained unchanged, while the correlation between the 2 groups was disrupted. The bivariate and partial correlations (with the presence of the other 7 predictors, as previously described) between ASI, PCS or FPQIII, and AUC were evaluated, and the difference in the absolute correlation values of ASI-AUC, PCS-AUC, or FPQIII-AUC was calculated. The P value for the observed difference in the correlation was obtained by comparing the difference in the absolute value of correlation with the values obtained from the permutated samples. The analysis was implemented using the statistics software R (http://www.r-project.org) and IBM SPSS version 20 (Armonk, NY).
Fifty patients were enrolled in this prospective study, and 39 women received epidural analgesia and underwent vaginal delivery. All 39 patients completed the study protocol and had recorded values for their preinduction questionnaires and response variables. Data from all 39 patients were analyzed. There were no patients lost to follow-up and no patients withdrew from the study. Table 1 describes the demographic and obstetric data of the study cohort. Mean ± SD scores for the ASI, FPQ, and PCS questionnaire of the study cohort were 73 ± 17, 16 ± 9, and 16 ± 9, respectively. The median (IQR) for self-rated anxiety, confidence, and analgesia expectations were 60 (20–75), 80 (25–90), and 6 (5–7), respectively.
The median time from onset of labor to epidural request was 295 (190–455) minutes. The median (IQR) duration of labor was 676 (451–847) minutes. No significant study-related adverse events were reported during the study period. Median (IQR) maternal satisfaction with labor analgesia was 92 (80–100).
All bivariate correlations between predictive questionnaires and outcome response measures (as well as the corresponding 95% CIs after controlling for false discovery rate) are reported in Tables 2 and 3. Although some pairs of response measures and predictive tests were significantly correlated as judged by raw P values unadjusted for multiple testing, none of these correlations reached statistical significance after adjustment. These variables were still included in the multivariate linear regression model to evaluate their combined ability to predict the outcome response measures.
Table 4 shows results from the multiple linear regression analysis of clinical response outcome measures and the prelabor questionnaires. Area under the labor pain × time curve (AUC) and fitted values from the multiple regression modeling is illustrated in Figure 1. The robustness of the regression models identified in Table 4 was assessed using a resampling approach (Tables 5 and 6). Of note, variables included in the final models for the response measurements using the full data set had the highest importance of count among all predictors, except anxiety 0 to 100 (count of 303). Correlations (as well as the corresponding 95% CIs) among the various psychological questionnaires are reported in Tables 7 and 8. There was no significant correlation between ASI and anxiety 0 to 100 (r = 0.03, P = 0.91).
The difference between ASI-AUC correlation and FPQIII-AUC correlation was significant (P = 0.027); that between ASI and PCS, as judged from a permutation test, was not (P = 0.067). Neither ASI (partial r = 0.31, P = 0.082, 95% CI: [−0.05, 0.60]), FPQIII (partial r = −0.12, P = 0.54, 95% CI: [−0.46, 0.25]), nor PCS (partial r = −0.15, P = 0.41, 95% CI: [−0.49, 0.22]) had a significant partial correlation with labor pain AUC when the other 7 potential predictors (EPQR-SF personalities of psychoticism, extroversion, neuroticism, lying, and anxiety, confidence and analgesic expectation rating) were considered. The difference between ASI-AUC and FPQIII-AUC or PCS-AUC was not statistically significant (P value = 0.11 and 0.17, respectively). When all other 7 predictors were included in the model without model selection, adding ASI, FPQIII, or PCS did not show significant additional contributions as judged by an F test (P = 0.09, 0.54, 0.42, respectively). After proper model selection, neither FPQIII nor PCS was in the final model, although ASI was still included in the final model (P = 0.022).
Our results suggest that ASI may contribute to the modeling of labor pain × time AUC and epidural local anesthetic use. Although ASI was included in the final multivariate linear regression model and FPQ and PCS were not, further study is required to determine whether ASI is a better predictor than FPQ or PCS. Previous studies have demonstrated that anxiety negatively impacts pain experienced during labor.13,14,27,28 The bivariate correlation between ASI and labor pain AUC of 0.44 was stronger than that reported by Reading and Cox,13 who found that although anxiety scores emerged as the best predictor of labor pain ratings, it only accounted for a further 5% of the labor pain variance. In the current study, we used ASI to measure a patient’s anxiety level, although the Spielberger State-Trait Anxiety Inventory (STAI) is more commonly used in clinical trials.3,27,29,30 We selected ASI because it is a validated questionnaire that is quicker to administer than STAI (and therefore more suited to the dynamic labor setting) and its efficacy compares favorably with STAI.31 The utility of STAI for pain prediction in an obstetric setting has been modest in studies that have used it.30
Among the Eysenck personality traits (psychoticism, extroversion, neuroticism, and lying)21,22 examined, lying contributed to the modeling of time to request for epidural analgesia, and extroversion and psychoticism contributed to the predictive modeling of labor pain × time AUC. There are limited data regarding the impact of personality and labor pain prediction. Eysenck22 reported that extraverts report significantly more discomfort during childbirth, whereas Reading and Cox13 found limited value in predicting labor pain with personality assessments. Data from the current study suggest that different personality traits may influence labor analgesia outcome measures differently: for example, extroversion and psychoticism relating more to labor pain and lying to the timing of epidural analgesia request.
Pain catastrophizing contributed to the multivariate linear regression analysis of the clinical outcome response of epidural local anesthetic use. Previous authors have found that patients with higher PCS scores anticipated and experienced more intense pain in childbirth,32 have increased requests for labor pain relief,33 and report higher pain on the second day after cesarean delivery.15 Van den Bussche et al.34 found that pain catastrophizing was not related to requests for epidural analgesia during labor. In the current study, although FPQIII did not predict labor pain or analgesic outcomes, women who expressed more fear of pain had lower maternal satisfaction with labor ratings. Fear of childbirth has been associated with likelihood of requiring unplanned cesarean delivery during labor.35 Differences in the utility of PCS and FPQ to predict the labor pain experience among studies may be explained by differences in patient demographic (level of education, parenting classes), obstetric (spontaneous versus induction of labor), and analgesic (epidural versus nonmedicated labor) characteristics.
Self-reported ratings of anxiety, confidence, and analgesic expectations showed some potential as predictive tests for the labor pain experience. Self-reported confidence and anxiety rating contributed to the multivariate linear regression modeling of labor pain × time AUC and epidural local anesthetic use, and anxiety and analgesic expectations contributed to the predictive modeling of time to epidural analgesia request. A previous study that investigated the relationships between the perception of pain during active labor and 9 predictor variables (age, parity, childbirth preparation, state anxiety, confidence in ability to handle labor, concern regarding the outcome of labor, fear of pain, cervical dilation, and frequency of uterine contractions) found that confidence in ability to handle labor was the most significant predictor of pain during active labor.36 However, the current study specifically assessed confidence to cope with labor pain and did not specifically focus on ability to handle labor. We found no significant correlation between ASI and self-reported anxiety 0 to 100, suggesting that anxiety assessments require a more complex assessment tool beyond a global rating scale.
There were a number of strong correlations among the various psychological questionnaires. In particular, FPQ correlated significantly with PCS, and PCS correlated significantly with psychoticism. The results suggest redundancy among the psychological questionnaires and indicate that psychological factors (such as anxiety, fear of pain, catastrophizing, and psychoticism) are related constructs. Pain catastrophizing has been found to be positively associated with the fear of being overwhelmed by labor pain.34 Although fear of pain and catastrophizing are related constructs, predictive values for pain rating outcomes among psychological factors do vary.5
Overall, our findings suggest that the psychological (ASI, FPQ, PCS) questionnaires, Eysenck personality assessments (psychoticism, extroversion, neuroticism, and lying), and other psychosocial evaluations completed before the onset of labor may help predict the labor pain experience. Bivariate correlations between prelabor psychological questionnaires and clinical response outcome measures varied from 0.01 to 0.44, and multiple regression analysis adjusted R 2 coefficient ranged from 0.27 to 0.45. The findings suggest that these psychological factors are generally less predictive of labor pain compared with postoperative pain.3–10,29 Potential explanations for why these psychological evaluations were of modest predictive value most likely relate to the predictors selected, the clinical setting, the outcome measures, and that psychological characteristics only contribute modestly to observed differences in labor pain. Although we measured a number of predictive psychological factors, we acknowledge that some psychological factors that affect labor pain were not accessed.1,2 We selected anxiety, fear of pain, and pain catastrophizing because these factors have been found to be associated with pain in a surgical setting. Although the psychological questionnaires are validated, they are neither labor-specific nor have they been validated in the labor setting. We did not assess mood because we felt that a depression score done just before the induction of labor would not accurately reflect the women’s baseline mood. In addition, while measures of mood, such as Profile of Mood States and Beck Depression Inventory, are important in chronic pain,37 their relevance for acute, dynamic pain in a labor setting is uncertain. Postoperative depression has been found to be associated with pain intensity, disability, and dissatisfaction after surgery,38,39 as well as the risk of persistent postpartum pain.40 Chronic sleeping difficulties have also been found to increase the risk of severe postoperative pain41; however, sleep quality was not assessed in the current study.
The labor pain experience is a very difficult clinical outcome to evaluate, with no ideal pain outcome variable to assess the experience.42 We measured a number of outcome measures including pain, time to epidural analgesia request, hourly epidural local anesthetic use, and maternal satisfaction. However, we acknowledge that these outcome measures may not completely represent or reflect the labor pain experience. We appreciate that there were a number of questionnaires to complete, and we did not measure respondent fatigue. For practical reasons, the study population included women undergoing induction of labor and focused on women who had vaginal delivery to produce a more homogeneous study sample and minimize potential confounders. We acknowledge that this may limit generalizability of the results, and future studies are required in this and other laboring populations, as well as different clinical settings. Although some pairs of response measures and predictive tests are significantly correlated as judged by raw unadjusted P values, none of these correlations reached statistical significance after adjustment for multiple testing. Therefore, the identified correlated pairs are only plausible and additional study is required to further assess the significance of the findings. The importance of counts of variables reported with the resampling analysis approach were not optimal. We believe that this was mainly due to the small sample size, and with additional samples, the results would be expected to be more significant and would likely appear more times in the resampling analysis.
In conclusion, ASI was a predictor of the labor pain experience; however, further study is required to determine whether ASI is a better predictor than FPQ or PCS. Personality traits (psychoticism, extroversion, and lying), as well as simple-to-administer, self-reported ratings of anxiety (0–100), confidence (0–100), and analgesia expectations (0–10), showed some potential to predict labor pain, epidural local anesthetic use, and time to epidural analgesia request. Future studies with larger sample sizes are required to better refine the ideal psychological measurement tools that can be administered before the onset of labor to predict the labor pain experience and to determine whether modifying labor analgesia methods based on these predictors will improve outcomes.
Name: Brendan Carvalho, MBBCh, FRCA.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Brendan Carvalho has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Ming Zheng, PhD.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Ming Zheng has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Leinani Aiono-Le Tagaloa, MBChB, FANZCA.
Contribution: This author helped design the study, conduct the study, and write the manuscript.
Attestation: Leinani Aiono-Le Tagaloa has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
This manuscript was handled by: Cynthia A. Wong, MD.
The authors thank Daniel Broderick for his contributions in helping set up the study and for initiating study enrollment and acknowledge the Foundation for Anesthesia Education and Research (FAER) for support of the Daniel Broderick research externship.
1. Lowe NK. The nature of labor pain. Am J Obstet Gynecol. 2002;186:S16–24
2. Hodnett ED. Pain and women’s satisfaction with the experience of childbirth: a systematic review. Am J Obstet Gynecol. 2002;186:S160–72
3. Ip HY, Abrishami A, Peng PW, Wong J, Chung F. Predictors of postoperative pain and analgesic consumption: a qualitative systematic review. Anesthesiology. 2009;111:657–77
4. Hinrichs-Rocker A, Schulz K, Järvinen I, Lefering R, Simanski C, Neugebauer EA. Psychosocial predictors and correlates for chronic post-surgical pain (CPSP)—a systematic review. Eur J Pain. 2009;13:719–30
5. Sullivan MJ, Thorn B, Rodgers W, Ward LC. Path model of psychological antecedents to pain experience: experimental and clinical findings. Clin J Pain. 2004;20:164–73
6. Granot M, Ferber SG. The roles of pain catastrophizing and anxiety in the prediction of postoperative pain intensity: a prospective study. Clin J Pain. 2005;21:439–45
7. Papaioannou M, Skapinakis P, Damigos D, Mavreas V, Broumas G, Palgimesi A. The role of catastrophizing in the prediction of postoperative pain. Pain Med. 2009;10:1452–9
8. Perry F, Parker RK, White PF, Clifford PA. Role of psychological factors in postoperative pain control and recovery with patient-controlled analgesia. Clin J Pain. 1994;10:57–63
9. Vaughn F, Wichowski H, Bosworth G. Does preoperative anxiety level predict postoperative pain? AORN J. 2007;85:589–604
10. Theunissen M, Peters ML, Bruce J, Gramke HF, Marcus MA. Preoperative anxiety and catastrophizing: a systematic review and meta-analysis of the association with chronic postsurgical pain. Clin J Pain. 2012;28:819–41
11. Tanum L, Malt UF. Personality and physical symptoms in nonpsychiatric patients with functional gastrointestinal disorder. J Psychosom Res. 2001;50:139–46
12. Asghari A, Nicholas MK. Personality and pain-related beliefs/coping strategies: a prospective study. Clin J Pain. 2006;22:10–8
13. Reading AE, Cox DN. Psychosocial predictors of labor pain. Pain. 1985;22:309–15
14. Lang AJ, Sorrell JT, Rodgers CS, Lebeck MM. Anxiety sensitivity as a predictor of labor pain. Eur J Pain. 2006;10:263–70
15. Strulov L, Zimmer EZ, Granot M, Tamir A, Jakobi P, Lowenstein L. Pain catastrophizing, response to experimental heat stimuli, and post-cesarean section pain. J Pain. 2007;8:273–9
16. Ferber SG, Granot M, Zimmer EZ. Catastrophizing labor pain compromises later maternity adjustments. Am J Obstet Gynecol. 2005;192:826–31
17. Reiss S, Peterson RA, Gursky DM, McNally RJ. Anxiety sensitivity, anxiety frequency and the prediction of fearfulness. Behav Res Ther. 1986;24:1–8
18. Asmundson GJ, Bovell CV, Carleton RN, McWilliams LA. The Fear of Pain Questionnaire-Short Form (FPQ-SF): factorial validity and psychometric properties. Pain. 2008;134:51–8
19. McNeil DW, Rainwater AJ 3rd. Development of the Fear of Pain Questionnaire–III. J Behav Med. 1998;21:389–410
20. Osman A, Barrios FX, Kopper BA, Hauptmann W, Jones J, O’Neill E. Factor structure, reliability, and validity of the Pain Catastrophizing Scale. J Behav Med. 1997;20:589–605
21. Eysenck SB. The validity of a personality questionnaire as determined by the method of nominated groups. Life Sci. 1962;1:13–8
22. Eysenck SB. Personality, and pain assessment in childbirth of married and unmarried mothers. J Ment Sci. 1961;107:417–30
23. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Statist Soc B. 1995;57:289–300
24. Fisher RA Statistical Methods for Research Workers. 197014th ed Davien, CT Hafner Publishing Company
25. Akaike H. A new look at the statistical model identification. IEEE Trans Autom Control. 1974;19:716–23
26. Royston P. An extension of Shapiro and Wilk’s W test for normality to large samples. Appl Stat. 1982;31:115–24
27. Beebe KR, Lee KA, Carrieri-Kohlman V, Humphreys J. The effects of childbirth self-efficacy and anxiety during pregnancy on prehospitalization labor. J Obstet Gynecol Neonatal Nurs. 2007;36:410–8
28. Wuitchik M, Hesson K, Bakal DA. Perinatal predictors of pain and distress during labor. Birth. 1990;17:186–91
29. Kalkman CJ, Visser K, Moen J, Bonsel GJ, Grobbee DE, Moons KG. Preoperative prediction of severe postoperative pain. Pain. 2003;105:415–23
30. Pan PH, Coghill R, Houle TT, Seid MH, Lindel WM, Parker RL, Washburn SA, Harris L, Eisenach JC. Multifactorial preoperative predictors for postcesarean section pain and analgesic requirement. Anesthesiology. 2006;104:417–25
31. Ochsner KN, Ludlow DH, Knierim K, Hanelin J, Ramachandran T, Glover GC, Mackey SC. Neural correlates of individual differences in pain-related fear and anxiety. Pain. 2006;120:69–77
32. Flink IK, Mroczek MZ, Sullivan MJ, Linton SJ. Pain in childbirth and postpartum recovery: the role of catastrophizing. Eur J Pain. 2009;13:312–6
33. Veringa I, Buitendijk S, de Miranda E, de Wolf S, Spinhoven P. Pain cognitions as predictors of the request for pain relief during the first stage of labor: a prospective study. J Psychosom Obstet Gynaecol. 2011;32:119–25
34. Van den Bussche E, Crombez G, Eccleston C, Sullivan MJ. Why women prefer epidural analgesia during childbirth: the role of beliefs about epidural analgesia and pain catastrophizing. Eur J Pain. 2007;11:275–82
35. Sydsjö G, Sydsjö A, Gunnervik C, Bladh M, Josefsson A. Obstetric outcome for women who received individualized treatment for fear of childbirth during pregnancy. Acta Obstet Gynecol Scand. 2012;91:44–9
36. Lowe NK. Explaining the pain of active labor: the importance of maternal confidence. Res Nurs Health. 1989;12:237–45
37. Dworkin RH, Turk DC, Farrar JT, Haythornthwaite JA, Jensen MP, Katz NP, Kerns RD, Stucki G, Allen RR, Bellamy N, Carr DB, Chandler J, Cowan P, Dionne R, Galer BS, Hertz S, Jadad AR, Kramer LD, Manning DC, Martin S, McCormick CG, McDermott MP, McGrath P, Quessy S, Rappaport BA, Robbins W, Robinson JP, Rothman M, Royal MA, Simon L, Stauffer JW, Stein W, Tollett J, Wernicke J, Witter JIMMPACT. . Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain. 2005;113:9–19
38. Seebach CL, Kirkhart M, Lating JM, Wegener ST, Song Y, Riley LH 3rd, Archer KR. Examining the role of positive and negative affect in recovery from spine surgery. Pain. 2012;153:518–25
39. Thomas T, Robinson C, Champion D, McKell M, Pell M. Prediction and assessment of the severity of post-operative pain and of satisfaction with management. Pain. 1998;75:177–85
40. Eisenach JC, Pan PH, Smiley R, Lavand’homme P, Landau R, Houle TT. Severity of acute pain after childbirth, but not type of delivery, predicts persistent pain and postpartum depression. Pain. 2008;140:87–94
41. Mamie C, Bernstein M, Morabia A, Klopfenstein CE, Sloutskis D, Forster A. Are there reliable predictors of postoperative pain? Acta Anaesthesiol Scand. 2004;48:234–42
© 2014 International Anesthesia Research Society
42. Carvalho B, Cohen SE. Measuring the labor pain experience: delivery still far off. Int J Obstet Anesth. 2013;22:6–9