The effect of aromatherapy with rose and lavender on anxiety, surgical site pain, and extubation time after open‐heart surgery: A double‐center randomized controlled trial

To determine the effect of aromatherapy with rose and lavender on the patient outcomes after open‐heart surgery (OHS). In the clinical trial, patients were randomized to four groups. One group received routine care, the placebo group received a cotton swab soaked in water and the other two groups received either a cotton swab containing three drops of rose or lavender essence (0.2 ml). A total of 160 patients were randomized into four groups. Intergroup anxiety was not significantly different; however, the reciprocal time‐group effect was significant among the four groups. The extubation time was significant among the four groups which related to rose essence group compared with the control group (p < .001) and placebo group (p = .029). The surgical site pain was significant in the rose essence and lavender groups compared to the control group. Aromatherapy can reduce extubation time, surgical site pain severity, and anxiety in patients undergoing OHS

Advancements in innovative treatments like thrombolytic treatment, balloon angioplasty, laser therapy, and arterectomy have improved the medical management of cardiovascular patients (Persell et al., 2020). However, cardiac surgery (CS) is still the mainline treatment for many patients with CVDs. The coronary artery bypass graft (CABG) surgery has been performed for about 35 years throughout the world (Rocha, 2017). CABG is used for the management of 26.79% in North America, 0.72% in Asia, 17.94% in Western Europe, and 18.14% in other parts of the world (Mary, 2017) CVD patients.
Also, 60% of open-heart surgeries (OHSs) in Iran are by CABG surgery . Following CABG surgery, the patients are managed in cardiac care unit (CCU) and mechanical ventilation is one of the most common postoperative treatments for all of these patients (Barzanji, Zareiyan, Nezamzadeh, & Mazhari, 2019;. The process of weaning the patients from mechanical ventilation is of utmost importance for CABG patients (Elgebaly & Sabry, 2018) since the long-term use of mechanical ventilation may not only increase the healthcare costs, but also may result in various adverse outcomes (Borges et al., 2017;Burns et al., 2019).
Moreover, physiologic responses to CABG will lead to stress, anxiety, and restlessness in CS patients predisposing to insomnia, increased oxygen consumption by myocardium, increased workload of the sympathetic system, tachypnea, tachycardia, neurohormonal responses, and hypertension, all of which make the extubation difficult (Hetland et al., 2017). Consequently, most CS patients receive sedatives and analgesics to diminish pain, stress, and anxiety and to facilitate mechanical ventilation (Arroliga et al., 2005). The pharmacological management with sedatives and analgesics are often accompanied by complications such as hypotension, suppressed vital functions such as respiration and heart rate, drowsiness, nausea, vomiting, constipation, drug dependence and tolerance, delayed extubation (Lee, Chung, Chan, & Chan, 2005), anaphylactic reaction, and even shock. In addition to numerous physical and mental complications, they also impose high healthcare cost to the economy .
Non-pharmaceutical methods have been increasingly recognized in recent years to reduce the extubation time and stabilize the vital signs . These non-pharmaceutical procedures such as music therapy  and educational interventions (Lee et al., 2018) offers a relatively simple, cheap, noninvasive, and low-complication option compared to the pharmaceutical medications (Avazah, Khosh Fetrat, & Rahimi Bashar, 2019).
Aromatherapy is one of the non-pharmaceutical option to manage perioperative patients. It is classified as a branch of alternative or complementary medicine. It involves the use of volatile herbal oils like essences to improve the psychosomatic health. It is used to relieve pain, anxiety, depression, insomnia, fatigue and exhaustion, asthma, and even self-confidence, success, and creativity (Hur, Song, Lee, & Lee, 2014). These oils can be used through inhalation, bath, or massage. The most common method used in aromatherapy is massage (Hur et al., 2014).
It is believed that the aroma activates the olfactory neurons leading to stimulation of the limbic system; as a result, various neural transmitters are released depending on the type of aroma. These transmitters include enkephalin, endorphin, noradrenaline, and serotonin (Hur et al., 2014). One of the aromas with a tranquilizing effect is lavender with the scientific name Lavandula stoechas (Lavandula angustifolia) of the Labiatae family (mints). Linalool and linalyl acetate are the active components of this herb. Linalool acts as a tranquilizer by affecting gamma amino butyric acid (GABA) receptors in the central nervous system (CNS) and linalyl acetate exerts a narcotic effect.
There are a few studies on aromatherapy in OHS patients such as cryotherapy and inhalation of rose aroma in reducing pain during extubation (Hasanzadeh et al., 2016), rose aroma on diminishing sternotomy site pain and the reduced need for analgesics (Heidari Gorji et al., 2015), rose aroma inhalation on stress and vital signs (Bikmoradi et al., 2015), and rose aroma on vital signs (Salamati, Mashouf, & Mojab, 2017). To the best of our knowledge, no study has so far compared the efficacy of rose and lavender aromas in reducing surgical site pain, anxiety and extubation time. Hence, this study aimed at determining the effect of aromatherapy with rose and lavender aromas on reducing extubation time, surgical site pain and anxiety in patients undergoing OHS.

| Design
The study was a double-center randomized clinical trial.

| Participants
The study was conducted in the CCU of Baqiatallah-e A'zam Hospital and Jamaran Hospital in Tehran, capital of Iran, in 2016. Patients were eligible for study participation if they had the following qualifications:

| Randomization
Convenience sampling method was used first, followed by random assignment with octad blocks in four groups (two intervention groups, one control group, and one placebo group). To perform the random assignment with octad blocks, the letter A was allocated to lavender group, B to rose group, C to placebo group, and D to control group.
Using the letters A, B, C, and D in octad groups, eight possible combinations of AABBCCDD, BBAACCDD, ABABCCDD, BABACCDD, ABBACDDC, BAABDCCD, AABBDCDC, and CDDCBAAB were recorded on separate sheets and cast in a box. One of these sheets was drawn out of the box randomly, the letters combination on it was written down again, and was recast in the box. Since the sample size was 180 in this study, this lottery was repeated 23 times and each time the new letters combination and the previous letters combination were noted. Then, a number ranging from 1 to 180 was allocated sequentially to each letter in the successive combinations. Subsequently, each letter was put in an opaque envelope on which numbers 1-180 were written. Each time a patient was selected, one of these pockets was opened on the basis of the number on it and this determined the group to which the patient ought to be assigned. The essence used for the intervention groups was volatile and could penetrate in the surrounding space interfering with other aromas. If both intervention groups were drawn out simultaneously in the lottery in the same day, the manipulation for the intervention groups was performed in separate hospitals to avoid the interference of aromas between the two groups.

| Intervention
In intervention groups, having obtained the permission of the university authorities, the researchers presented to Baqiatallah-A'zam Hospital and Jamaran Hospital in Tehran and explained the research goals and procedures to the head nurse in charge. The samples who qualified for inclusion as participants were selected in the hospital setting the day before surgery. Informed written consent was obtained from each patient. The demographic and clinical information questionnaire was completed by the patients and in some cases by the use of their medical records. At the completion of surgery, the patients were admitted to the CCUs. After transferring to the CCU, the first basic vital signs were recorded from the monitor for the four groups before dripping rose and lavender essence or the placebo immediately after the first triggered inspiration by the ventilated patient or the spontaneous respiration by the patient. The main components of rose and lavender oils were citronellol (41.23%) and ceineole (21.33%), respectively. After triggering of the first inspiration, a cotton swab soaked in three drops of rose or lavender essence (0.2 ml) was placed on the patient's chest for 15 min. Then, the vital signs were recorded halfhourly before extubation and quarterly after extubation for 1 hr. The frequency of vital signs recording differed from patient to patient depending on their extubation time. The patient's anxiety level was measured preoperatively and immediately after entrance to the operation room using Spielberger anxiety questionnaire. It was also reexamined postoperatively and after transference to the CCU with the patients being awake. The surgical site pain after wheeling into CCU was investigated with a visual analogue scale while the patient is awake.
In placebo group, the procedure was exactly similar to that of the intervention groups except that a cotton swab soaked in water was placed on the patient's chest. In control group, the procedure was exactly similar to that of the intervention and placebo groups except that this group received just routine care with no intervention.

| Outcomes and relevant measures
The primary outcome was determining anxiety and the secondary outcomes were surgical site pain and extubation time after OHS.

| Physiological indices sheet
The sheet included systolic pressure, diastolic pressure, heart rate, respiration rate, and arterial blood oxygen saturation at various time intervals. These indices were measured by vital signs monitoring system (Dotex, General Electric Company, USA). The cardio respiratory and pulse oximetry monitoring device was calibrated first by the medical engineering of the center on the basis of manufacturer's guidelines and its accurate functioning was ascertained. The extubation time was measured with a standard chronometer (Saba Iran Co., Iran).

| The visual analog scale
The visual analog scale (VAS) was used to measure surgical site pain.
This scale is a 10-cm ruler ranging from 0 to 10 cm. In this scale, 0 represents no pain and 10 indicate intolerable pain Chen, Huang, Wu, Kuo, & Lin, 2019). The reliability coefficient of this tool was estimated as ICC = 0.91 using test-retest reliability coefficient and Cronbach's α = .93.

| Spielberger standard 20-item questionnaire
This questionnaire was used to examine anxiety. This inventory uses a four-point Likert scale with a score ranging from 4 to 80. Higher scores indicate more severe anxiety (Vahedian-Azimi et al., 2016). The reliability coefficient of the anxiety questionnaire was estimated as ICC = 0.92 using test-retest method and Cronbach's α = .94.

| Sample size estimation
The sample size was estimated to be 36 patients in each group using Altman nomogram with Type I error of 0.05, test power of 90%, and effect size of 0.9 on the basis of Kavehee et al.'s study (Kavei, Ebadi, Saeed, Moradian, & Sedigh Rahimabadi, 2015). Considering a subject attrition rate of 10%, 40 patients were assigned to each group adding up to 160 patients in all.

| Statistical analysis
All analyses were conducted using SPSS 23 (SPSS Inc., Chicago, IL) and GraphPad Prism 5© (Graph Pad Software Inc., La Jolla, CA) at α = .05 significance level. Data were presented using mean (SD) for normal numeric variables and frequency (Percentage)  160 were included in the final analysis. Forty patients under CABG surgery were excluded due to emergency CS (n = 6), CS more than one time (n = 3), intra-aortic balloon pump after surgery (n = 3), coagulating problems (n = 5), and declined to participate (n = 3; Figure 1).

| Nonadjusted and adjusted analyses of states anxiety
The results of two-way ANOVA with repeated measures on anxiety indicated a significant interaction between time and groups in the unadjusted and adjusted analyses representing a different time dependency in anxiety among groups in before and after measurements, after adjusting for age, gender, marital status, qualification, having history of diabetes, pulmonary disease, smoking, hypertension, hospitalization period, and BMI. Unadjusted and adjusted between group comparisons by utilizing ANOVA, did not showed significant differences in state anxiety in the rose, lavender, and placebo groups compared to the control group (p > .05). Additionally, the results of within group comparisons did not showed significant differences in state anxiety in all groups (p > .05; Table 2).

| Nonadjusted and adjusted ordinal logistic regression of pain
The results of unadjusted ordinal logistic regression showed significant lower pain severity in the rose and lavender groups compared to the placebo and control groups. The lowest pain severity was seen in the rose and lavender, respectively; although the severity did not show significant differences between the placebo and control but the severity was more in the placebo than control group. In the adjusted analyses, the pain severity was significantly lower in the rose and the lavender groups compared to the placebo and control group after adjusting for age, gender, marital status, qualification, diabetic, and smoking history; The lowest pain severity was seen in the rose and lavender, respectively; although the severity did not show significant differences between the placebo and control but the severity was more in the placebo group than control (Table 3 and Figure 2).

| Nonadjusted and adjusted analyses of extubation time
The results of one way ANOVA showed a significant difference among groups (p> .0001). Also the results of Tukey post hoc tests indicated significant differences between the rose and placebo groups (p = .029) and between the rose and the control groups (p < .0001). Also the results of ANCOVA after adjusting for pump time, graft number, ejection fraction, age, smoking, respiratory disease, and pulse rate, respiratory rate, systolic and diastolic blood pressure, and SPO 2 in 13 times showed significant differences among the groups (p < .0001). Also the results of Tukey post hoc tests indicated significant differences between the rose and placebo groups (p = .041) and between the rose and the control groups (p = .012; Table 4 and Figure 3). There was a significant difference in extubation time among the groups in the adjusted and nonadjusted analyses that pertained to the rose group compared with control and placebo groups. The results showed that demographic information (age, gender, BMI, education, marital status, and smoking) and underlying diseases (a history of hypertension, diabetes, repeated hospitalization, and pulmonary and coagulation problems) were the same among the four groups before intervention while variables pertaining to OHS (cardiac pump time, VEF, and number of grafts) were the same after intervention among the four groups (Table 1).

| DISCUSSION
The results of surgical site pain in this study was similar to the study by Heidari Gorji et al. (2015), but different from the results of Salamati et al. study that showed that the inhalation of lavender essence exerted no significant effect on decreasing pain after OHS. This difference may be attributed to method, type, and time of intervention or the mean age of the patients. Salamati et al.'s study included one group with a mean age of 50 years in which aromatherapy was done for 10 min (Salamati, Mashouf, Sahbaei, & Mojab, 2014).
For anxiety, the results of Babaii, Abbasinia, Hejazi, Seyyed Tabaei, and Dehghani (2015) and Muzzarelli, Force, and Sebold (2006) studies were similar to the pairwise intergroup comparisons of this study. The study by Babayee et al. did not compare the reciprocal time-group effect of intervention; nevertheless, this comparison was significant in our study; in other words, time changes among the groups were not the same after intervention and were significantly different (Babaii et al., 2015). Moreover, the results of Wilkinson et al. revealed that the patients who received aromatherapy massage had improved anxiety and depression levels compared to those who received merely ordinary care during 10 weeks. This is consistent with the intergroup results of this study; however, it has been effective 2 weeks after intervention which is consistent with the reciprocal time-group effect of this study (Wilkinson et al., 2007). Our findings indicated that aromatherapy with essences of rose and lavender induced a significant effect on reducing extubation time.
The extubation time was reduced in the rose group in both adjusted and nonadjusted analyses compared to the control and placebo groups. Since early extubation is done in almost all cardiovascular surgery centers round the globe due to its numerous advantages, and given that reduced extubation time results in greater rehabilitation and earlier discharge from CCUs that prevents the numerous investigating the effect of reflexive plantar massage on physiologic indices and extubation time in OHS patients, showed no effect on these indices; yet, it exerted an effect on reducing extubation time, a finding which is consistent with this study (Ebadi, Kavei, Moradian, & Saeid, 2015). Apart from the promising effects of aromatherapy in these patients, other factors including medication adherence , health literacy , and self-care behaviors (Lin et al., , 2020Saffari et al., 2017) in the patients with CVDs have protective effects.
There are certain limitations for this study. The researchers tried their best to reduce the effects of their presence on the participants (observer paradox). These included any uncontrollable environmental noises in the ward that could influence the aromatherapy process, the physical, mental, and personality structure of the individuals which were not identifiable and could cause differences in the effect of aromatherapy on patients, though the patients were randomly assigned to the four groups. The extubation process was commenced by the CCU nurse in charge, there were some differences in nurses' performance that could cause differences in the results, though an orientation session was held to homogenize the nurses' performance in the intended wards. Completing some tasks such as taking X-rays and moving the patients sometimes awakened them leading to the administration of narcotics.

| CONCLUSION
Aromatherapy with rose and lavender resulted in a significant difference in anxiety in the reciprocal time-group effect in the adjusted and nonadjusted analyses among the groups under study. The extubation time was significantly different among the groups in the adjusted and nonadjusted analyses which pertained to the rose group compared to the control and placebo groups. The severity of pain was less in the two intervention groups compared to the control and placebo groups, though albeit it was less in the control group than in the placebo group. It could be possibly asserted that aromatherapy is a useful nursing intervention in reducing pain severity and extubation time after CS. It is a cheap and complication-free procedure that can promote healthcare level and increase patient convenience at the critical time after OHS. Also, given the contradictory results of the related studies, it is recommended that more extensive studies be carried out in future on the effect of aromatherapy on patients' anxiety level. According to the analysis of covariance (ANCOVA) adjusted for pump time, graft number, ejection fraction, age, smoking, respiratory disease, and pulse rate, respiratory rate, systolic and diastolic blood pressure, and SPO 2 in 13 times.
F I G U R E 3 Extubation time in four groups lavender, rose, placebo, and control