Scale (m-YPAS): Modified Yale Preoperative Anxiety; Scale (STAIC): State and Trait Anxiety Inventory for Children; Scale (STAI): State and Trait Anxiety Inventory; Scale(CAPS-Anxiety): Children Anxiety and Pain Scales; CSWQ: Child Surgery Worries Questionnaire; Scale (SAM): Self-assessment Mannequin Scale; EAS: Child’s Temperament; Scale (FAS): Facial Affective Scale; PBH: Posthospital Behavior Questionnaire; CCPH: Cuestionario de Conducta Post-hospitalaria; Scale FLACC: Face, Legs, Activity, Cry and Consolability; Scale (CHEOPS): Children’s Hospital of Eastern Ontario Pain Scale; Score(VAS): Visual Analog; PB: Blood Pressure; EMLA: Local Anesthetic Cream (EMLA®, Astrazeneca, London, UK); CL: Child Life

Pediatric patients are stressed by separations from their parents, strange environments, fears of painful treatments and also treatment outcome insurances [1,2]. There are several ways to reduce anxiety and pain other than premedication. Among them, laughter and humor were reported as an effective way to reduce pain, increase stress tolerance during medical evaluation and painful procedures [3,4] and were developed by several methods as clown intervention distraction. Hospital clowns use distraction methods such as magic, music and storytelling to enable children to deal with emotions (such as fear, anxiety, and boredom) they may experience while there are in the hospital.

Hospital clowns (or clown doctors) were reported to be beneficial for the pediatric patients with these stressors and to circumvent the accompanying feelings of fear, helplessness and sadness, thus supporting the healing process; they help to make a difficult situation easier, more bearable or simply offer a welcome distraction from the ward routine [4,5]. Hospital clowning was developed in 1986 in the United States by Michael Christensen from the New York-based Big Apple Circus, and spread quickly from there throughout Europe. In 2011 the European federation of hospital clowns was founded [6].

Previous reviews reported the effects of clown hospital on anxiety and pain but there was no comprehensive review of studies methodical design which assessed benefits of hospital clowning interventions [3,7,8]. The aim of this literature review is to analyze design of the studies, relating to the benefit of hospital clown interventions for hospitalized children.

A structured literature search was conducted via PubMed, National Library of Medicine. We have included all publications focusing on clowning hospital interventions for children under 18 years in the pediatrics services that were published in English, in French, and in Spanish from 2005 to 2016. The literature search involved a combination of key words including clowning OR clown OR clown intervention OR therapeutic clowning OR clown AND children OR clown doctors OR hospital clowning OR clown OR medical clowns OR medical clowning OR clown therapy OR clown anxiety.

We scanned titles and abstracts and retrieved the full text of potentially relevant articles, scanned the reference lists. Disagreements were resolved by consensus discussion. The articles were then independently reviewed by 2 of the authors. For each publication included, we analyzed: population targeted, medical specialties, medical procedure, anxiety and pain measurement instruments, other clinical outcomes and data focused on methodological points: clinical trial design, definition of target population, randomization, singleblind assessment, sample size calculation and statistical analysis. We have also reported the benefit of clown intervention on anxiety and/or pain or other clinical outcomes medical procedure. Any discrepancies were resolved by consensus with a third reviewer.

The search via PubMed in February 2016 produced a total of 188 potential publications. We selected 54 of them on the basis of the relevance of the title. After reading their abstracts we selected 28 publications (Figure 1) [4,5,9-34]. The majority of studies (23 studies) were published between 2010 and 2016. These studies were published in different countries (Table 1).

Figure 1: PRISMA 2009 flow diagram of review process and study selection.

Table 1: Characteristic of studies included.

General characteristics of studies

The main characteristics of the 28 studies are presented in Table 2. The majority of studies (n=18, 64%) are randomized prospective and controlled studies [5,10-26]. One third of the studies targeted specifically the benefits of the intervention of hospital clown for children [17,19-22,27,29,30]. Over half of the studies reported intervention benefits of hospital clown for children and their parents [5,11-12,14,15,18,23-26,28,31-34]. Four other studies have evaluated in addition the opinions of health professionals [4,10,13,16]. One study targeted only mother's anxiety [9].

Table 2: General analyses of studies (studies design, population targeted, medical service evaluation, medical procedure, outcome evaluation, statistical analysis).

The target population studies varied from a study to another. The age group most highly represented was 2 to 12 year-old (43%) [9-19,21]. The other inclusion criteria of studies were children in good health (children without a history of chronic illness, non-premature birth, non-developmental delay and non-significant hearing or visual impairments), children who had no experience of previous anesthetic and/or first attempt examination with or without sedation and the need to undergo painful procedures.

Thirty two percent of studies were focused on anesthesia before surgery as minor surgery, strabismus surgery, otolaryngology surgery [9-14,18,21-22]. Thirty six percent of studies were focused on invasive medical procedures including venous blood sampling, intravenous catheter, burn, wound dressing, immobilization of injured limbs, wound suture and blood tests [15-16,23], medical forensic examiner (allegedly abused children) [5,24-25], allergy prick skin tests 28, botulinum toxin injections 30, intra-articular corticosteroid injection 20, magnetic resonance imaging 17. Thirty two percent of studies targeted non-invasive procedures [4,19,26,27,29,31-34].

General conditions for the performances of hospital clowns

The main characteristics of the performances of hospital clowns are presented in Table 3. The clown intervention started at the waiting or medical Preoperative room and/or Emergency department. Working methods of clowns may vary across studies, as their professional training. Over half of studies of clown intervention lasted less than 15 minutes [3,4,12,15,16,21,22,25,27-34]. For seventy one percent of the studies the intervention is realized by one clown at one visit [4,5,9,11,15,16,19,21,23-33] and in twenty nine percent of the studies the clowns intervention is realized by two clown at one visit [10,12-14,17,18,22,34].

Table 3: General Conditions for the performances of hospital clowns (Methods distraction, Time of intervention, Number of clown).

Studies design

Among the 28 included studies, 18 are randomized prospective and controlled studies and 14 of them have parallels groups design (clowngroup/ control-group) [9-13,16,17,21-23,28-30,34]. Half of the studies have a sample size of less than 30 patients per group [9,11,13,14,16,18,21,23,29]. In 2 of 28 studies, the evaluators were blinded to both, the study’s design, hypothesis [18] and patients or parents group assignment [19].

Variables of outcomes and instruments

The main tools of efficacy measure are presented in Table 4. Most of the studies measured the impact of the clown intervention on anxiety and/or pain. In most of the cases, anxiety and pain were measured with standardized and validated self-report or observer-rated questionnaires, specifically developed for children according to the age group [6,10-13,16-26], or by visual analog scales [15,20,23,28] or crying duration [15,30]. Other clinical outcomes medical procedure were the length of time spent in hospital 29 or the systolic blood pressure [11,20,29]. In most of the cases, the well-being was evaluated by interview [31-33].

Table 4: Tools of efficacy mensuration (Anxiety, Pain mensuration, Crying duration, Clinical assessment, Well-being).

Statistical analyses methods

Only 8 studies presented the sample size calculation [9,11,12,14-16,19,30]. To determine the sample size the child’s and/or parent’s anxiety as primary endpoint was mainly used [9,11,12,14-16,19]. Only one study used the crying duration as primary endpoint 30. For comparison of anxiety, pain perception and/or wellbeing between the experimental group and control group, 12 of the 18 randomized prospective studies used univariate analysis [11-16,19,21-23,28-29], 6 studies used multivariate analysis [14,15,18,21,23-24] and only one cross sectional study without parallel groups used multivariate analysis 27 (Table 2).

Efficacy of clown intervention

Figure 2 describes efficacy of clown intervention for the 28 studies according to their different characteristics in design, using reduction in anxiety or pain as primary end point.

Figure 2: Diagram describing efficacy of clown’s intervention for the 28 studies.

Anxiety (fear, stress, worries) level reduction

Twenty-one studies evaluated clown hospital intervention on the anxiety of children and/or parent, including one study on parent’s anxiety, 8 studies on children’s anxiety and 12 studies on both parent’s and children’s anxiety. We observed that 10 of 20 studies reported a positive impact on reducing children’s anxiety [10-17,27,28] and 7 of 13 studies reported a positive impact on reducing parent’s anxiety [10-12,19,23,24,26]. Hence, 50% of the total reviewed studies (n=28) presented a positive reduction of parent’s or children’s anxiety or pain, regardless of their study design.

Pain level reduction

Among 28 studies, there are 7 randomized prospective and controlled studies that evaluated impact effects of clown intervention on the children’s pain perception during invasive medical procedure and/or post-operative period [11,15,16,20,23,28,29]; 5 of 7 studies showed a positive impact with significantly lower level of pain perception in the clown group compared with the control group [11,15,16,28,29]. However, none of them have calculated the sample size with the pain as the primary endpoint. None of the studies used multivariate analysis; hence no contingent confounding factor was taken into account.

Clinical assessment

Only one randomized prospective and controlled study has evaluated the impact of the presence of clown intervention on clinical outcome 29, but did not reveal any impact on the length of hospitalization, duration of the fever period and time taken to achieve clinical recovery. However, this study did not report any sample size calculation, which could argue for a non-significant result consequent to power analysis lacking.

Mental well-being and magical attachment

We identified 4 studies which examined the impact of clown interventions on well-being [31-34] of children and/or their parents. Three of them were qualitative studies without parallel groups and quantitative measures [31-33] and one of them was randomized prospective and controlled study. In this last study, authors used a multivariate analysis and reported that clown intervention had no impact on the well-being of children and their parents [34].

Opinion of health professionals regarding the presence of clowns in the preoperative or medical procedure room

The opinion of the health professionals is shown in Table 5. We identified 4 studies that evaluated the opinion of health professionals regarding the presence of clowns [4,10,13,16]. The validated opinions of health professionals are different from a study to another. This review showed that 44% to 93% of health professionals were satisfied with clown interventions and believed that the clown program was beneficial to the hospital and children’s health care management [4,10,13,16]. Two studies showed that the majority (70% to 92%) of the health professionals opposed to continue the clown intervention, because of perceived interference with procedures in the preoperative or medical room (64%) [13,16]. However, one study showed that 96% of health professionals agreed with the presence of clowns continuing in the preoperative room, mainly because they considered that clowns were not disturbing agents (71.43%) [10]. These review show that health professionals considered clown interventions useful for the children (75% to 96.43%), for the parents (40% to 89.29%) and for themselves (35% to 64.29%) [4,10,13,16].

Table 5: Assessment report on the medical staff of the interaction of the clowns.

This literature review brings a comprehensive, systematic review of methodical design of studies which assess benefits of interventions of clowning hospital, and shows the variety of methods used to assess the impact of intervention of hospital clowns.

The results of our research reveals that only half of the studies concluded in favor of the intervention of the hospital’s clowns, meanwhile the second half was against the intervention. This literature review was not able to show us the conclusions on the benefit of the clown intervention.

Among 28 studies analyzed, if the majority of studies (n=18) are randomized prospective and controlled studies, few of them have high level of evidence: only 8 of the 18 prospective studies complied the calculation of the sample size to compare the groups with and without intervention and reported the use of a primary endpoint [9,11,12,14-16,19,30], among them 2 studies had single-blind assessments. Only 7 of the studies used multivariate analysis [14,15,18,21,23,24,28], 6 of them were randomized prospective and controlled studies and took into account contingent confounding factors and the last one was cross-sectional.

Also, this literature review shows the variety of measurement tools, which makes it difficult to compare these studies. Most of the authors used tools that do not allow single-blind assessments. As it is an observer-rated questionnaire, only the m-YAS permitted to assess anxiety with blinded evaluators to group type.

Among 28 studies shown, clown interventions are heterogeneous in a study to another. The majority of studies used the same methods of distraction, such as magic tricks, music, funny stories, funny songs, telling jokes and puppets. Over half of study reported that the clown’s intervention took less than 15 minutes in one intervention [3,4,8,12,15,16,21,22,25,27-33]. Again, it is of the utmost importance to get more homogeneity in the type of intervention and duration to improve the knowledge and the efficacy of the clown’s intervention.

These results underlie the need for a better methodological assessment in future studies with stronger evidence-based approaches to strengthen positive effects of clowning interventions. Weaknesses in design methods and multiple assessment tools utilizations highlight the heterogeneity in measures. This may contribute to the methodological issues that future studies should consider to produce more robust, comparable and accurate estimates.

None of the studies are taking into account medico-economic evaluation of clown intervention. The clown intervention is very expensive with the documentation of the hourly wage (with an average rate of € 43.00 per hour) but previously non-existent transparency has been achieved 6. Moreover, clown intervention may also increase the total length of care and also the total cost, thus leading to question about its advantages in terms of cost-benefit. These points urge for future studies to develop more medico-economic evaluation (such as cost minimization studies, cost-effectiveness, cost-utility and costbenefit analyses), and indicate the methodological problems commonly encountered in this field.

This review provided some empirical evidences of the effect of clown interventions, using laughing and humor, on anxiety and/or pain reduction in pediatric hospitals. Indeed, for some studies assessed, hospital clowns contributed to minimize children's and parent’s anxiety levels, decrease children's surgery-invasive medical procedure-related pain and showed positive effects of psychological well-being in children’s hospitals. However, the findings of the present study also suggested that conditions of the clown interventions should be defined better and diversity of the studied samples and measurement tools should be addressed, hence demonstrating that more evidence-based studies are needed in the future in the field of hospital clowns interventions.

The authors declare that they have no competing interests.

SP: 1st author, involved in data collection, in drafting the manuscript; ST: 2nd author, involved in revising the manuscript; MA: 3rd author, involved in revising the manuscript; SG: last author and corresponding author involved in revising the manuscript, approved the final version to be published. All authors read and approved the final manuscript for this submission.

We thank Popescu D. and Stefan O. for the English revision.