Remember me
Acute pain is a frequent occurrence in hospital care and particularly in the emergency department (ED), due to the presenting condition as well as the investigative or therapeutic procedures. Optimizing its management in this setting is a constant endeavor (1). Although pharmacological management is standard of care, enhancing nonspecific aspects of treatment is also recommended (2). Nurses are in general especially aware of nonspecific aspects of care and use them (3). Communication is a key component that can elicit possible placebo or nocebo effects, especially during critical time points of care, such as procedural information (4,5). Positive suggestions can be introduced into this communication and have the potential to alleviate acute pain perception (6–8).
Using positive verbal communication or suggestions is rich in promises for this clinical context. However, in a number of prior studies with positive outcomes, research team members delivering the suggestions were aware of the hypotheses and their behavior was not monitored (9–11), which could have led to biases and misattribution of the effects. Indeed, suggestions are a verbal part of communication, yet nonverbal aspects are important contributors to nonspecific treatment effects (12). In fact, a recent study of positive suggestions conducted with stringent methods shows limited effects and reveals the complexities of the clinical setting (13).
This study was designed to test whether positive suggestions compared with usual communication could attenuate pain and anxiety perceived during the placement of a peripheral intravenous catheter (PIC) in the ED, and if these suggestions needed to be made by the nurse in charge of the procedure to be effective. To this aim, a modulation of the communication was implemented: patients who underwent PIC placement were randomly attributed to a positive versus standard (negative) message. A first phase of the study delivered the message through an audio recording via a headset, and the nurse was asked to remain silent during the procedure. In the second phase, a trained nurse who was performing the procedure delivered the message. It was predicted that the nurses would have a more empathic behavior when delivering the positive than the standard message or when silent. The hypotheses were that a positive message provided through an audio recording would lead to less pain and anxiety than the usual one, and it would have an enhanced analgesic and anxiolytic effect if delivered by a person in charge of care (with empathic behavior).
METHODS Study DesignThis was a randomized, controlled, sequential group, monocentric trial with a 2 + 2 design: positive versus standard message; phase 1: audio taped–phase 2: nurse delivered. The first phase was double-blind, testing a positive versus standard information provided through a headset. The nurse performing the PIC placement was asked to have minimal interaction with patients during the procedure. The second phase was single-blind, performed by three nurses trained to deliver both intervention messages. The sequential design was chosen to guarantee that the large ED clinical nursing team would not be confused as to their role in the study. The first phase took place between October 2017 and October 2018, followed by the second phase in October 2019, at a teaching hospital ED with ~45,000 consultations a year. The study was approved by the Cantonal Ethics Committee for Vaud (CER-VD-2017-01505) and registered on ClincalTrials.gov (NCT03502655).
ParticipantsAll patients 18 years or older who were admitted to the ED for a nonvital emergency and clinically stable, were requiring a PIC on the upper limb, able to interact in French, and give informed consent were invited to participate on the active study days. All participants gave an oral consent to the study before participation, based on partial information that omitted the randomization to different communication contents, to prevent expectation effects. Full written informed consent was obtained after collecting all outcomes. Participants were free to retract their participation at any stage.
ProcedureEligible patients were approached in the ED by a study team member and given an information sheet, which described an observational study aiming to improve care in the ED (no mention of communication or randomization). After preliminary oral consent, randomization was performed by opening sequentially sealed, numbered opaque envelopes, which contained a code based on a computer-generated random 1:1 allocation (variable block sizes of two, four, and six; separate coding system for each study phase). The opening date was noted on the envelope, and the opening sequence was regularly checked by the lead investigator.
The standard message was based on usual communication observed in the study ED (pilot phase where warnings of pain, clinical terms, etc., were transcribed) consistent with previously published findings (14) (“You may feel the tourniquet placed on your arm, it gets tight, maybe it tingles. The sting of the inserted needle can be painful, but don’t move at all during the procedure. When this message stops, raise a finger with the other hand, and the caregiver will sting you.”).
The positive message was based on the literature and positive communication scripts (15) (“Your arm may begin to feel a certain numbness, like a kind of heaviness, or warmth, as if it were asleep for the duration of the procedure. Now that you’ve heard this message, as soon as you’re ready, you can signal this to the caregiver by raising a finger with the other hand.”).
Phase 1The randomization code corresponded to the numerical identity of prerecorded messages held on an MP3 reader. Both the study team members and the nursing team were blind to the message. The patient was provided with a Bluetooth-connected headset, and when the nurse prepared for the PIC (disinfection etc.), the prerecorded message was played. Patients who revealed contents of the message to the study team or to whom the nurse talked between preparation of PIC placement and outcome measures were excluded.
Phase 2The randomization envelope indicated which message (positive or standard) was to be delivered by the nurse (single-blind). While preparing and inserting the PIC, the nurse delivered the allocated message. A study team member checked compliance with the attributed message (requirement of 4/5 elements underlined in the scripted message and no contradictory elements to maintain inclusion of patient).
After the observation phase of the PIC placement and study-related outcome collection, complete written information regarding the study was provided, and written consent was collected.
Measures Main OutcomesAnxiety and pain intensity were collected through self-reported visual analog scales (VAS) with anchors 0 = not at all and 100 = extremely, after PIC placement.
Secondary OutcomesSecondary outcomes were pain and anxiety VAS at baseline (pre-PIC placement). The study team observed a) nonverbal expressions of pain (Algoplus scale) (16) and b) empathetic behavior in the patient-nurse interaction: behavior enhancing the clinical encounter according to placebo research using a purpose-built five-point scale (1 point for each behavior: “warm and friendly manner,” “active listening,” “verbal empathy,” “thoughtful silence during 20 seconds,” and “communication of confidence or positive expectations”) (17).
Data Analysis Primary Outcome AnalysesThe study investigated the efficacy of the interventions, and we used a per-protocol analysis. A sample of 55 patients per group was calculated as necessary to detect clinically relevant changes in pain VAS in the ED (18), with a power of 80% and an α error at .05. Considering attrition of about 10%, inclusion of 65 patients per group was planned. Missing data were not imputed.
Patients’ characteristics were compared between the four arms using two-tailed Pearson χ2 tests for discrete variables and Kruskal-Wallis tests for continuous variables. All tests were two-tailed. Given the randomized design, to assess the average intervention effect, the outcome distributions were compared between the two study arms, separately for each phase, using the Wilcoxon-Mann-Whitney test. Given the 2 + 2 design, no statistical comparisons of outcomes were conducted between phases.
Exploratory AnalysisIn a recent publication on the effect of 3D virtual reality on procedural pain, we showed a potential heterogeneous differential analgesic and anxiolytic effect of virtual reality based on preprocedural anxiety level (19). The randomized clinical trial presented here was also designed to assess the average effect of positive communication on procedural pain and anxiety. Because randomized clinical trials are not designed to assess the heterogeneity of treatment effects across individuals (20), we conducted post hoc exploratory analyses using linear multivariate regression analyses to detect a possible heterogeneous effect of the intervention (i.e., positive versus standard communication) (21,22).
The analyses were performed with STATA, version 16.1 (StataCorp, College Station, Texas) by a research collaborator blinded to the study groups.
The data that support the findings of this study are openly available in Zenodo at http://zenodo.org (reference number 10.5281/zenodo.7534606).
RESULTSThe patient flowchart is presented in Figure S1, Supplemental Digital Content, https://links.lww.com/PSYMED/A966. The baseline characteristics (Table 1) show that randomization in each phase was successful, with well-matched characteristics. There were significant differences in the population’s characteristics between study phases, with patients in the second phase less likely to have a surgical condition, chronic pain and to be chronically on analgesics.
TABLE 1 - Patient Baseline Characteristics, Showing the Control and Intervention Groups for Both Study Phases Phase 1IQR = interquartile range; VAS = visual analog scale; PIC = peripheral intravenous catheter.
p Values are presented for the comparison between the two phases. Statistics: two-tailed Pearson χ2 tests for discrete variables and Kruskal-Wallis tests for continuous variables.
There was a significant difference between the nurses’ empathic behavior score when delivering the positive versus control message, interpreted as a positive manipulation check: as expected, the empathic behavior was less in phase 1 than 2, because nurses were asked to communicate minimally with patients in phase 1, whereas in phase 2, they were more empathic in the positive than the standard message group (Table 2).
TABLE 2 - Main Study Outcomes by Phase and Intervention Group Phase 1VAS = visual analog scale; PIC = peripheral intravenous catheter; IQR = interquartile range.
Statistics: two-tailed Pearson χ2 tests for discrete variables and Kruskal-Wallis tests for continuous variables.
There were no statistically significant differences in patients’ reported pain and anxiety during PIC insertion between communication groups, whether in phase 1 or 2 (Table 2). There was no effect of the message on pain behavior (as measured with Algoplus).
In addition, patients in phase 2 were more likely than in phase 1 to need two attempts at catheter placement (phase 2: n = 28; phase 1: n = 1). There was a trend toward more catheter placement attempts in the standard message group compared with the intervention in phase 2 (Table 2).
The heterogeneity analysis, controlling for different patient characteristics, did not show significant effects of our positive communication during either phase on pain intensity or pain unpleasantness during PIC placement (Supplemental Digital Content, Tables S1–S4, https://links.lww.com/PSYMED/A966). There was a borderline significant effect of the communication (β coefficient = 9.2 mm, 95% confidence interval = 0.3 to 18.1, p = .044) on anxiety during PIC placement in phase 2 (Supplemental Digital Content, Table S6), but not in phase 1 (β coefficient = 1.6, 95% confidence interval = −6.8 to 10, p = .7; Supplemental Digital Content, Table S5). Patients with baseline anxiety greater than 30/100 mm, with pain 50/100 mm and greater, and having trypanophobia (fear of needles) had greater procedural anxiety, accounting for the intervention (Supplemental Digital Content, Table S6). The multivariable linear regression analyses showed some other interesting findings regarding heterogeneity of intervention effect. Severe preprocedural pain contributed to an increased pain intensity during the procedure, but not pain unpleasantness. Trypanophobia significantly increased pain intensity, pain unpleasantness, and anxiety during both phases and types of communication. Interestingly, preprocedural anxiety did not impact procedural pain intensity or unpleasantness. Preprocedural anxiety was associated with greater procedural anxiety. Nurses’ empathy had a different impact during phase 1 and 2. During phase 1, when nurses were minimally interactive and responsive, greater empathy tended to be associated with lower procedural pain intensity and unpleasantness, whereas the opposite was found when nurses interacted with patients in phase 2. Nevertheless, given the lack of significance of most of the models, they cannot be interpreted as evidence that the intervention had a positive effect on certain patients. This conclusion would require multiple interaction analyses, which our study was not powered to conduct. The details of these analyses are presented in Tables S1 to S6, Supplemental Digital Content, https://links.lww.com/PSYMED/A966.
DISCUSSIONIn this real-life study in a tertiary hospital ED, a positive message was delivered through a prerecorded audio or by the nurse in charge of the procedure. Contrary to our hypothesis, this positive communication did not result in lower levels of procedural pain or anxiety during a PIC placement, compared with a standard message. No effects were obtained whether this positive message was delivered through audio recordings or trained nurses despite higher observed empathy scores in nurses delivering a positive message.
We will first discuss these negative results, with different hypotheses and suggestions for further work. The power question comes first. The study was carefully set up, with a conservative effect size calculation based on clinical relevance in the given context (18) and prior studies showing moderate effect sizes of verbal suggestions on pain (23). To assess if certain patient characteristics lead to the null result, a heterogeneity analysis was conducted. When controlling for the baseline factors in the heterogeneity analysis, no effect of the intervention could be shown on pain in either phase. Nevertheless, about 25% of the patients declared having trypanophobia, and they reported significantly more pain. This study did not have sufficient power to conduct multiple interaction analyses examining potentially significant effects of the intervention specifically in this subgroup. Based on our findings, future research targeting patients with trypanophobia could be warranted.
The second question concerns the choice and delivery of the intervention messages. The messages were simple scripts taught in a 1-hour training. Our positive message was based on prior works (9–11) and corresponds to positive communication training recommendations (14,15). Hence, our findings raise the possibility that the positive effect of communication in prior studies was due to the delivery of the intervention by unblinded study staff (messenger effect) or other elements, rather than the scripted verbal message. The training was successful, without any exclusion due to a failure of message delivery.
Third, there could have been a floor effect on pain scores. The pain reported from the PIC procedure (i.e., <20/100 on average) was lower than reported in similar studies (9,11). An earlier study with positive suggestions regarding PIC also reported very low pain scores on average and limited results, that is, no analgesia but less frequent pain-related vocalizations (10). Furthermore, our study patients had on average moderate discomfort due to their underlying condition. Hence, the pain induced by the PIC procedure could have been perceived as relatively minor. Therefore, such an intervention on communication might be better suited for a patient population without underlying pain and/or undergoing a procedure that is more painful. Interestingly, a recent study of hypnotic communication during PIC insertion in the operating room in pain-free patients showed higher PIC pain ratings (mean in neutral group = 3.5/10), with positive effects of the intervention (mean pain in the hypnosis group = 1.5/10) (24). In addition to the absence of underlying pain, the anesthesiologists delivering the messages in that study were trained in hypnosis, a technique having shown reliable effects in the procedural context (25). The interactions between preprocedural pain, intensity of noxious input, and communication interventions with different levels of complexity deserve to be further tested.
Beyond the negative main results, two findings deserve a brief discussion. The heterogeneity analysis on anxiety in phase 2 showed a statistically significant effect of the positive message. In other words, when controlling for all baseline factors, the intervention increased anxiety by 9.2/100 mm, a statistically significant result but a clinically insignificant result (26). In addition, this effect was marginal compared with the impact of preprocedural anxiety: that is, highly anxious participants had higher anxiety by 62 points on the 100-mm VAS than those without anxiety. Future efforts in the ED should be focused on anxious patients, who might need specific interventions. In fact, we also showed that preprocedural anxiety >12 mm blocked the analgesic effect of virtual reality during minor procedures (19). Finally, when delivering standard messages including usual nocebo suggestions in that context (14), nurses tended to have more PIC failures than during a positive message/not having to deliver a message. This unexpected finding of a potential self-delivered nocebo effect of a standard, negatively laden message, on nurses trained in positive communication is worthy of further research.
One of the main limitations of our study is the sequential 2 + 2 rather than a 2 × 2 design. This was chosen to guarantee that there was no cross-contamination between the different nursing roles (silent role versus message delivery) given our setting with a large nursing staff and high turnover of personnel.
In conclusion, the positive communication intervention did not impact the pain and anxiety reports after a peripheral catheter placement in the ED, whether the communication was audio-taped or delivered by a trained nurse. There might have been a floor effect, with low procedural pain ratings in a context of moderate pain due to the presenting condition. Hence, such a simple therapeutic communication intervention might not be sufficient, especially to modulate a mildly painful input in the ED.
The study team thanks Marie-France Derkenne for her help with monitoring the study and all the patients who agreed to participate.
Author Contributions: C.B. and O.H. planned the study, supervised data collection, led the data interpretation, and wrote the manuscript. O.H. and P.T. led the analyses. A.B., A.F.B., N.G., H.G.-D. contributed to study planning, enrolled patients, and collected the data. A.G. contributed to study planning and data interpretation. All authors critically revised the manuscript.
Source of Funding and Conflicts of Interest: A.G. is part of the board of the Institute for Medical hypnosis for the French-speaking part of Switzerland. The authors have no other conflicts of interest to declare. C.B. was supported by the Foundation Mercier Pour la Science during the time of writing. The Foundation had no role in the study presented here.
REFERENCES 1. Optimizing the treatment of acute pain in the emergency department. Ann Emerg Med 2017;70:446–8. 2. Colloca L, Barsky AJ. Placebo and nocebo effects. N Engl J Med 2020;382:554–61. 3. Palese A, Cadorin L, Testa M, Geri T, Colloca L, Rossettini G. Contextual factors triggering placebo and nocebo effects in nursing practice: findings from a national cross-sectional study. J Clin Nurs 2019;28:1966–78. 4. Alfano M. Placebo effects and informed consent. Am J Bioeth 2015;15:3–12. 5. Krauss BS. “This may hurt”: predictions in procedural disclosure may do harm. BMJ 2015;350:h649. 6. Koban L, Jepma M, Geuter S, Wager TD. What’s in a word? How instructions, suggestions, and social information change pain and emotion. Neurosci Biobehav Rev 2017;81:29–42. 7. Peerdeman KJ, van Laarhoven AIM, Donders ART, Hopman MTE, Peters ML, Evers AWM. Inducing expectations for health: effects of verbal suggestion and imagery on pain, itch, and fatigue as indicators of physical sensitivity. PloS One 2015;10:e0139563. 8. Mistiaen P, van Osch M, van Vliet L, Howick J, Bishop FL, Di Blasi Z, et al. The effect of patient-practitioner communication on pain: a systematic review. Eur J Pain 2016;20:675–88. 9. Varelmann D, Pancaro C, Cappiello EC, Camann WR. Nocebo-induced hyperalgesia during local anesthetic injection. Anesth Analg 2010;110:868–70. 10. Dutt-Gupta J, Bown T, Cyna AM. Effect of communication on pain during intravenous cannulation: a randomized controlled trial. Br J Anaesth 2007;99:871–5. 11. Ott J, Aust S, Nouri K, Promberger R. An everyday phrase may harm your patients: the influence of negative words on pain during venous blood sampling. Clin J Pain 2012;28:324–8. 12. Kaptchuk TJ, Friedlander E, Kelley JM, Sanchez MN, Kokkotou E, Singer JP, et al. Placebos without deception: a randomized controlled trial in irritable bowel syndrome. PloS One 2010;5:e15591. 13. van Vliet LM, Godfried MB, van Deelen GW, Kaunang M, Kaptchuk TJ, van Dulmen S, et al. Placebo effects of nurses’ communication alongside standard medical care on pain and other outcomes: a randomized controlled trial in clinical tonsillectomy care. Psychother Psychosom 2020;89:56–8. 14. Lang EV, Hatsiopoulou O, Koch T, Berbaum K, Lutgendorf S, Kettenmann E, et al. Can words hurt? Patient-provider interactions during invasive procedures. Pain 2005;114:303–9. 15. Lang EV. A better patient experience through better communication. J Radiol Nurs 2012;31:114–9. 16. Rat P, Jouve E, Pickering G, Donnarel L, Nguyen L, Michel M, et al. Validation of an acute pain-behavior scale for older persons with inability to communicate verbally: Algoplus. Eur J Pain 2011;15:198.e1–198.e10. 17. Kaptchuk TJ, Kelley JM, Conboy LA, Davis RB, Kerr CE, Jacobson EE, et al. Components of placebo effect: randomised controlled trial in patients with irritable bowel syndrome. BMJ 2008;336:999–1003. 18. Gallagher EJ, Liebman M, Bijur PE. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001;38:633–8. 19. Bosso L, Espejo T, Taffe P, Caillet-Bois D, Christen T, Berna C, et al. Analgesic and anxiolytic effects of virtual reality during minor procedures in an emergency department: a randomized controlled study. Ann Emerg Med 2023;81:84–94. 20. Angus DC, Chang CH. Heterogeneity of treatment effect: estimating how the effects of interventions vary across individuals. JAMA 2021;326:2312–3. 21. Burke JF, Hayward RA, Nelson JP, Kent DM. Using internally developed risk models to assess heterogeneity in treatment effects in clinical trials. Circ Cardiovasc Qual Outcomes 2014;7:163–9. 22. Kent DM, Rothwell PM, Ioannidis JPA, Altman DG, Hayward RA. Assessing and reporting heterogeneity in treatment effects in clinical trials: a proposal. Trials 2010;11:85. 23. Peerdeman KJ, van Laarhoven AIM, Keij SM, Vase L, Rovers MM, Peters ML, et al. Relieving patients’ pain with expectation interventions: a meta-analysis. Pain 2016;157:1179–91. 24. Fusco N, Bernard F, Roelants F, Watremez C, Musellec H, Laviolle B, et al. Hypnosis and communication reduce pain and anxiety in peripheral intravenous cannulation: Effect of Language and Confusion on Pain During Peripheral Intravenous Catheterization (KTHYPE), a multicentre randomised trial. Br J Anaesth 2020;124:292–8. 25. Lutgendorf SK, Lang EV, Berbaum KS, Russell D, Berbaum ML, Logan H, et al. Effects of age on responsiveness to adjunct hypnotic analgesia during invasive medical procedures. Psychosom Med 2007;69:191–9. 26. Williams VS, Morlock RJ, Feltner D. Psychometric evaluation of a visual analog scale for the assessment of anxiety. Health Qual Life Outcomes 2010;8:57.
Comments (0)