Exploring the Effectiveness of External Use of Bach Flower Remedies on Carpal Tunnel Syndrome

Abstract

Background.

A randomized, pilot, placebo-controlled clinical trial was conducted with the aim of evaluating the effectiveness of a cream based on Bach flower remedies (BFR) on symptoms and signs of carpal tunnel syndrome.

Methods.

Forty-three patients with mild to moderate carpal tunnel syndrome during their “waiting” time for surgical option were randomized into 3 parallel groups: Placebo (n = 14), blinded BFR (n = 16), and nonblinded BFR (n = 13). These groups were treated during 21 days with topical placebo or a cream based on BFR.

Results.

Significant improvements were observed on self-reported symptom severity and pain intensity favorable to BFR groups with large effect sizes (η² _partial > 0.40). In addition, all signs observed during the clinical exam showed significant improvements among the groups as well as symptoms of pain, night pain, and tingling, also with large effect sizes (φ > 0.5). Finally, there were significant differences between the blinded and nonblinded BFR groups for signs and pain registered in clinical exam but not in self-reports.

Conclusion.

The proposed BFR cream could be an effective intervention in the management of mild and moderate carpal tunnel syndrome, reducing the severity symptoms and providing pain relief.

Bach flower remedies (BFR) is a complementary therapy that deserves further scientific investigation. Edward Bach was a physician who used highly diluted preparations mainly from many derivatives species of wildflowers in order to help individuals recover their health. This topic had been very controversial, but it could be supported by the new knowledge about nanoparticles and its effects on living systems^1–4 rather than some “unknown healing energy.”

Although BFR has showed potentialities for pain management,⁵ there is no evidence of its specific action beyond the placebo effect.^6–8 However, recent data suggest that some BFR could have specific effects on inflammation,⁹ cardiovascular risk factors,¹⁰ spiritual well-being,¹¹ and unwanted intrusive thoughts.¹²

In clinical practice, classical selection of the remedy is mostly guided by individuals’ mood or their personality traits,¹³ but beyond this individualized treatment, there are anecdotic experiences of pain relief among patients with carpal tunnel syndrome (CTS) using a cream based on a BFR combination.¹⁴ Placebo analgesia pathways are well documented,^15–17 and it could result in support for the use of BFR as an ethical self-help placebo.^18–20

However, uncertainty about treatment allocation in randomized clinical trials could affect both treatment and placebo response.^21,22 On the other hand, in clinical practice patients tend to believe that they receive an “active” treatment, even when they are using a placebo intervention.^21,23 As can be suggested, belief is an amazing healing device,²⁴ which acts on behavior as a self-fulfilling prophecy.²⁵ As it has been stated, the contexts involved in randomized clinical trials and clinical practice are quite different.²⁶ Because of this, our study included a third group in which patients received the intervention as is usual in clinical practice.

CTS is a frequent entrapment neuropathy.^27–30 Primary features of CTS include pain in the hand, unpleasant tingling, pain or numbness in the distal distribution of the median nerve (thumb, index, middle finger, and the radial side of the ring finger), and a reduction of grip strength and function of the affected hand.²⁷ Surgical and nonsurgical treatments have been suggested for CTS.^31,32 Surgery is usually considered for patients with an experience of conservative treatment failure and those who have severe CTS, while nonsurgical treatments are usually prescribed as an initial option for the patients who suffer from nonconstant symptoms of mild to moderate CTS.³²

Multiple alternative nonsurgical techniques have been trialed.³⁰ Complementary and alternative medicine can play an important role as new conservative treatments for CTS.^33–35 Researchers suggest new formats to manage CTS, in order to improve cost-effectiveness using topic treatment applications.^35,36 The topical use of BFR might be a useful resource to many individuals with CTS with the purpose of achieving pain relief. Compared with other alternatives, it is cheap, easy to apply, and entails only a minimal contact with the patients.

The aim of the current pilot study was to evaluate the effectiveness of a cream based on BFR on symptoms and signs of CTS. With the additional aim of assessing the magnitude of the effect caused by the certainness of receiving an intervention, we introduced a nonblinded BFR group as is usual in clinical practice.

Methods

Participants

The design consisted of pilot therapeutic interventions in 43 outpatients who were aged between 20 and 89 (mean = 50.90; standard deviation = 13.97), with 93% being females (n = 40). Each individual had been previously diagnosed with CTS (from mild to moderated degree), according to clinical and electrographic criteria. The patients were referred to the Orthopaedic and Rheumatology Services at University Hospital “Arnaldo Milián Castro” with signs and symptoms of CTS for more than 3 months duration. In those with bilateral symptoms, the arm with the most severe symptoms was chosen, and treatment of this arm was randomized.

Every patient had already been treated for symptoms of the CTS without improvements and they were referred to surgical option after been treated conservatively. Physicians invited patients to participate in this study during their “waiting” time for surgical option. The intervention involved outpatients who were randomly allocated into 3 groups: 2 blinded groups as is usual in double-blind controlled trials, which were Placebo (n = 14) and BFR (n = 16); and 1 nonblinded group, which also received BFR (n = 13) as is usual in clinical practice.

Patients with neurological symptoms and signs suggestive of widespread peripheral neuropathy (such as sensory symptoms in the lower limbs and depressed or absent tendon reflexes) were excluded from the study. The patients were not receiving treatment with non-steroidal anti-inflammatory drugs (NSAID), neither were they subdued to local treatments at least a week before.

The flow of participants through the experiment, including reasons for exclusion, is depicted in Figure 1.

Figure 1.

Consolidated Standards of Reporting Trials (CONSORT) diagram showing the flow of participants through the study. BFRb, blinded BFR; BFRnb, nonblinded BFR.

Measures

Physician’s Report of Signs and Symptoms

Clinical exam was employed in order to detect signs of Tinel, Durcan, and Phalen as well as a set of core symptoms (see Table 1). The physician reported the presence or absence of each sign and symptom using a binary code of classification (0 = No or Absence; 1 = Yes or Presence).

Table 1.

Report of Signs and Symptoms During the Baseline Period.

	Positive Reports		PLA	BFRb	BFRnb	χ2(2)	Sig.	φ
	Freq.	%	Presence/Absence
Pain	43	100	14/0	16/0	13/0	—	—	—
Night pain	42	97.7	13/1	16/0	13/0	2.121	.346	0.22
Pain after Exercise	39	90.7	14/0	12/4	13/0	7.444	.024	0.41
Pain Hand & Fingers	28	65.1	10/4	11/5	7/6	1.066	.587	0.15
Tingling sensation	42	97.7	13/1	16/0	13/0	2.121	.346	0.22
Numbness	18	41.9	2/12	11/5	11/2	14.878	.001	0.58
Burning	19	44.2	6/8	9/7	4/9	1.903	.386	0.21
Valleix	23	53.5	10/4	11/5	2/11	10.896	.004	0.53
Temperature sensitivity	19	44.2	7/7	10/6	2/11	6.471	.034	0.39
Swelling sensation	32	74.4	12/2	12/4	8/5	2.074	.355	0.20
Thenar atrophy	10	23.3	14/0	11/5	5/8	6.499	.039	0.38
Weakness	31	72.1	13/1	14/2	4/9	15.922	.000	0.61
Tinel	39	90.7	14/0	14/2	11/2	2.200	.333	0.22
Phalen	41	95.3	13/1	15/1	13/0	0.992	.631	0.14
Durcan	41	95.3	14/0	14/2	13/0	3.540	.170	0.28

Abbreviations: PLA, placebo group; BFR, Bach flower remedies; BFRb, blinded BFR group; BFRnb, nonblinded BFR group.

Patient’s Self-Report of Symptom Severity and Daily Pain Intensity

Symptom Severity Scale (SSS) was assessed using the 11 items corresponding to the 11-item subscale of symptom severity from the Boston Carpal Tunnel Syndrome Questionnaire,³⁷ which is a disease-specific questionnaire referring to a typical 24-hour period in the past 2 weeks.

Visual Analogue Scale of Pain Intensity (VAS) is a daily self-report based on the visual analogue scale.^38,39 The scale was presented as a 10-cm line, anchored by verbal descriptors that follow a Likert-type scale interval, 0 = “no pain” to 10 = “worst imaginable pain.” The patients were asked to report their pain intensity every day. Weekly means were calculated during the baseline period (1 week prior to treatment) and during the treatment since the first to third weeks.

Procedure

The current pilot study was approved by the Scientific Council of the University Hospital “Arnaldo Milián Castro” and its medical ethics committee (No. 211/2011) prior to starting the study. Patients gave written informed consent prior to assessment and intervention.

The study consisted of 2 basic stages: (1) Baseline (pre-intervention) and (2) postintervention (21 days after baseline).

1. Baseline: First, the objectives of the study were explained and consent forms were signed prior to assessment. A specialist in orthopedics and another specialist in rheumatology performed the clinical exam and diagnosed symptoms and signs. Three general practitioners were in charge to collect self-reports for both SSS and VAS. Then, patients were randomly allocated either to the placebo or the BFR group by their corresponding general practitioner and given their instructions (see below).

2. Postintervention: 21 days later, the participant returned to the physician’s office and the assigned general practitioner collected self-reports while both the orthopedic specialist and rheumatologist repeated the clinical exam.

BFR used in the current study were provided by Healing Herbs Ltd and prepared in the Homeopathic Pharmacy (Santa Clara, Villa Clara, Cuba), based on solid petrolatum. The personal of this pharmacy created a code to identify creams in order to blind both participants and researchers; the code was revealed to researchers once the project was over.

Five BFR (Ulmus procera [pain relief], Ornithogalum umbellatum [recovering after trauma], Verbena officinalis [reduce inflammation], Clematis vitalba [reduce numbness and tingling], and Carpinus betulus [strength recovery]) were combined into a single formula (100 μL of each remedy from the stock bottle per 100 mL of solid petrolatum). Solid petrolatum was used as the placebo control. Each cream was numbered and its content was unknown to physicians or patients, except for the nonblinded BFR group. All the containers were identical in shape, color, and size.

Those patients allocated to the blinded groups were told that they would receive either BFR or placebo during their “waitlist” period, whereas those patients allocated to the nonblinded group were told that the will receive a BFR cream. Patients were asked to apply the cream on the affected area every morning and night, over 21 days.

Statistical Analysis

Data analyses were performed using SPSS for Windows (version 20.0). A mixed ANOVA design (GLM for repeated measures) was used. The within-subject factor was “Time,” referring to the mean of the dependent variables throughout the baseline (BL) to the end of the “waiting” period (21 days). The between-subject factor was “Group,” referred to the blinded groups and the nonblinded group. The analysis of the interaction Time × Group was considered the main outcome analysis for the current study. Effect sizes were expressed with η² _partial whereby values of 0.01, 0.06, and 0.14 were defined as small, medium, and large, respectively.⁴⁰

Regarding the physician’s report of signs and symptoms, we conducted both global and specific analyses. Global analyses were performed creating 2 new variables named sum of symptoms and sum of signs (see Table 2), respectively, which were analyzed using the aforementioned mixed ANOVA.

Table 2.

Baseline Characteristics of the Groups and Change Over Time on Dependent Variables.

	PLA, Mean ± SD	BFRb, Mean ± SD	BFRnb, Mean ± SD	Statistics			Effect Size, η2 partial
				F	df	Sig.
SSS-BL	2.98 ± 0.50	3.07 ± 0.35	3.25 ± 0.29	21.771	2,40	.000	0.52
SSS-final	2.84 ± 0.50	1.69 ± 0.89	1.43 ± 0.59
∑Signs-BL	2.92 ± 0.26	2.68 ± 0.79	2.84 ± 0.37	21.681	2,40	.000	0.52
∑Signs-final	2.85 ± 0.53	1.37 ± 1.20	0.38 ± 0.86
∑Symptoms-BL	8.14 ± 1.46	8.56 ± 2.12	7.30 ± 1.25	7.901	2,40	.001	0.28
∑Symptoms-final	7.00 ± 1.88	3.75 ± 3.54	2.69 ± 1.97
VAS of pain BL	8.00 ± 1.83	8.43 ± 1.49	8.15 ± 0.89	13.762	6,120	.000	0.40
VAS of pain week 1	7.86 ± 1.85	7.17 ± 1.33	6.74 ± 1.27
VAS of pain week 2	7.77 ± 1.51	5.88 ± 1.44	5.36 ± 1.49
VAS of pain week 3	7.57 ± 1.75	3.95 ± 2.74	1.91 ± 2.48
	Recovered/Equal/Worsened			χ2	df	Sig.	φ
Pain	0/14/0	7/9/0	12/1/0	23.291	2	.000	0.73
Night pain	1/13/0	10/6/0	12/1/0	20.484	2	.000	0.69
Pain after exercise	2/12/0	6/10/0	0/13/0	6.916	2	.031	0.40
Pain hand and fingers	1/12/1	4/12/0	6/7/0	7.064	4	.133	0.13
Tingling sensation	1/13/1	11/5/0	10/3/0	20.642	4	.000	0.69
Numbness	1/13/0	3/13/0	6/7/0	6.038	2	.049	0.37
Burning	2/11/1	7/8/1	4/9/0	4.010	4	.405	0.35
Valleix	3/10/1	8/7/1	4/9/0	8.348	4	.080	0.44
Temperature sensitivity	3/10/1	8/8/0	1/12/0	8.832	4	.065	0.45
Swelling sensation	3/11/0	7/9/0	7/6/0	3.153	2	.207	0.27
Thenar atrophy	0/14/0	1/15/0	0/13/0	1.178	2	.422	0.20
Weakness	5/9/0	7/9/0	1/12/0	4.717	2	.095	0.33
Tinel	0/14/0	9/7/0	10/3/0	17.677	2	.000	0.64
Phalen	1/12/1	6/10/0	12/1/0	21.452	4	.000	0.76
Durcan	1/13/0	6/10/0	10/3/0	13.775	2	.001	0.56

Abbreviations: PLA, placebo group; SD, standard deviation; BFR, Bach flower remedies; BFRb, blinded BFR group; BFRnb, nonblinded BFR group; SSS, Symptom Severity Scale; VAS, Visual Analogue Scale; BL, baseline period; ∑Signs, sum of signs; ∑Symptoms, sum of symptoms.

Note: Bold values are those values with effect sizes higher than medium.

On the other hand, specific between-group analyses of signs and symptoms were performed using the χ² test. We additionally created a new variable named difference by subtracting values at the 21st day of treatment from baseline, with only 3 possible outcome values, 1, 0, and −1, which indicate worsening, no changes, and recovering, respectively. Effect sizes were calculated with φ, where values of 0.1, 0.3, and 0.5 were defined as small, medium, and large effects, respectively.⁴¹ All statistical tests are reported 2-tailed, and P values marked as ns refer to P > .05.

Results

Baseline Period

Groups did not differ in gender (χ²[2] = 2.140, P = .343), age, SSS, sum of signs, and sum of symptoms (F[2, 40] between 0.474 and 2.008, P > .005). As can be seen in Table 1, signs of Tinel, Phalen, and Durcan as well as a considerable number of symptoms did not show significant between-group differences.

However, significant differences with large effect sizes (φ > 0.5) were reported for pain, reported as dull or aching discomfort, in the hand, forearm, or upper arm; numbness; and weakness of hand and fingers. Numbness showed more positive reports among participants included in the BFR groups than those in the placebo group. On the other hand, participants included in the blinded groups reported more pain, reported as dull or aching discomfort, in the hand, forearm, or upper arm and weakness than those included in the nonblinded group.

Effectiveness of the Intervention

The effect of Time was significant for the all assessed variables (F_{SSS-∑Signs-∑Symptoms}[1, 40] between 68.990 and 110.595, P < .001, η² _partial > 0.60; F_VAS[3, 120] = 65.142, P < .001, η² _partial > 0.62). However, as can be derived from Table 2, researchers were more interested in the effect of Time × Group where there were significant differences favorable to BFR groups over placebo with very large effect sizes (η² _partial > 0.5).

The effect of Group brings support to the above-mentioned changes favorable to the BFR groups (F[2, 40] between 7.081 and 16.126, P < .001, η² _partial > 0.25). Table 3 displays the post hoc analyses in which BFR groups were grouped into the same subset of effects. It is important to highlight that the nonblinded BFR group differed from placebo for all analyses.

Table 3.

Results From Post Hoc Analyses: HSD Tukey.

	SSS		∑Sings		∑Symptoms		VAS of Pain
	S1	S2	S1	S2	S1	S2	S1	S2
BFRnb	2.34		1.61		5.00		5.54
BFRb	2.38		2.03		6.15	6.15	6.36
PLA		2.91		2.89		7.57		7.80
Sig.	.964	1.000	.167	1.000	.192	.089	.251	1.000

Abbreviations: PLA, placebo group; BFR, Bach flower remedies; BFRb, blinded BFR group; BFRnb, nonblinded BFR group; SSS, Symptom Severity Scale; VAS, Visual Analogue Scale; ∑Signs, sum of signs; ∑Symptoms, sum of symptoms.

Regarding the effectiveness of the intervention on each specific sign, results were also favorable to BFR groups as can be seen in Table 2. Furthermore, there were also significant differences between both BFR groups for Phalen with large effect size (χ²[1] = 9.151, P = .002, φ = 0.56) and for Durcan with medium effect size (χ²[1] = 4.507, P = .034, φ = 0.39).

Pain, night pain, and tingling were those symptoms in which a large number of individuals from BFR groups were reported as recovered with large effect sizes over the placebo group (Table 2). In addition, BFR groups differed on pain relief with a large effect size favorable to the nonblinded group (χ²(1) = 7.486, P = .006, φ = 0.51). Finally, although worsening of symptoms was scarcely reported among both blinded groups these reports were not observed among nonblinded individuals.

The observed differences between blinded and nonblinded BFR groups for signs of Phalen and Durcan and for symptom of pain were an interesting finding because these improvements were diagnosed by physicians. On the other hand, both blinded and nonblinded BFR groups tended to be grouped into the same subset of effect when patients’ self-reports of symptom severity and weekly mean values of the daily report of pain intensity were analyzed.

Finally, significantly less number of patients needed surgery among BFR groups (blinded n = 7 and nonblinded n = 1) when compared with the placebo group (n = 13) with a large effect size (χ²[2] = 19.831, P < .001, φ = 0.68). Furthermore, there was significant difference between both BFR groups favorable to the nonblinded group but with medium effect size (χ²[1] = 4.668, P = .031, φ = 0.40).

Discussion

The current research, performed with the aim of evaluating the effectiveness of a BFR cream on symptoms and signs of CTS showed interesting results that, although encouraging, should be discussed with caution.

The observed therapeutic outcomes that were favorable to the BFR groups over placebo in reducing signs, symptom severity, and pain as well as a less number of patients who needed surgery provide evidence that topical use of BFR might be an effective complementary treatment for CTS.

Multiples alternative nonsurgical treatments have been trailed in CTS. Oral steroids for a short period is efficacious after a month, but the long-term efficacy of oral steroid treatment remains uncertain.⁴² Steroid and procaine injections are effective in CTS regarding short-term (for 2 months) and long-term outcomes compared with placebo injections, reducing the symptoms of CTS.⁴³ However, the use of steroids is contraindicated for some patients.

With regard to the topical treatment for CTS, few studies were found. A clinical trial evaluating EMLA cream (lidocaine 2.5% plus prilocaine 2.5%) had some beneficial effects after 4 weeks of treatment.³⁶ Some studies evaluated herbal preparations in the management of CTS such as linseed oil.³⁵

The current results also support the idea that BFR may exert some biological effects,^9,10 specifically on numbness independent of whether a group was blinded or not. The controversial effects of high dilution could be clarified with the understanding of nanoparticles,⁴⁴ its features,⁴⁵ and the biological effect found in the field of homeopathy.^2,46 It is almost well established that there are certain differences between those complementary and alternative medicine modalities, but some researchers point to the existence of derivate from biological synthesis of metallic nanoparticles in plant extracts, bacteria, algae, and fungus,⁴⁷ and the rapid sunlight-induced efficient biogenic synthesis of nanoparticles⁴⁸ could be due to the effect of the “sun method” in the preparation process of BFR.⁴⁹ It might mean that the nonexistent particles in BFR would be just a misconception.

It was interesting to note that the placebo group did not show substantial improvements taking into account that the placebo effect on pain⁵⁰ or hand surgery⁵¹ as well other symptoms of CTS has been reported in previous studies. Possibly the context of the study itself could explain the lack of placebo effect.²⁶ In our study, we offered an alternative treatment while patients were waiting for the surgical option (the “real” treatment). Maybe the perception of treatment as “something to do” in the meanwhile and not as the main option impaired the mechanism of placebo response. However, this idea should be tested in future studies.

Results among nonblinded patients also deserve to be analyzed because they reported the greatest outcomes because significant differences observed with blinded BFR on signs and pain could be mainly attributed to the clinical exam rather than patients’ self-reports. This finding suggests that doctors involved tended to overestimate the effect of the treatment when they believe they have prescribe the right treatment.

In fact, a previous study using BFR provides indicators that users tended to underestimate its effects. Hyland and colleagues found that after a self-help treatment with BFR only 54% of the participants perceived improvements but the rates increased to 83% when they reported symptoms improvements.²⁰

In addition, results of the nonblinded group support the idea that regular clinical practice (in which uncertainty because allocation does not exist) exerts a context effect that tends to improve the healing properties of the assessed interventions.⁵²

The study’s contribution should be considered in light of its main limitation, which was the small sample size. However, sample size apparently seems an obstacle when results favor BFR over placebo. For example, the study of Pintov and colleagues,⁵³ which reported 12 and 11 children in the BFR and placebo group, respectively, has been cited in recent systematic reviews as an evidence of nonspecific effects.^6–8 On the other hand, positive results against placebo from an interesting subgroup into the study of Halberstein and colleagues,⁴⁹ with 22 and 17 participants in the BFR and placebo groups, respectively, are scarcely mentioned.

As it has been stated, many valuable studies might have been missed due to the publication bias known in this area.⁵⁴ Fortunately, there is new evidence that supports the idea that BFR may have specific effects.^9–12

This study was also limited due to consideration of short-term effects. Further research is recommended to determine the long-term outcomes and comparison of BFR creams with other conservative managements of CTS.

To conclude, the BFR cream could be an effective intervention in the management of mild and moderate carpal tunnel syndrome, reducing the severity symptoms and providing pain relief.