Predictors and Use of Nonpharmacologic Interventions for Procedural Pain Associated with Turning among Hospitalized Adults

Materials and Methods 

Results 

Sample 

This report focuses on one specific turn for each of the 1,395 adult patients in the Thunder II study who were turned. Table 1 reports the demographic and clinical characteristics of these patients. Patients were predominantly white (86.3%) and male (55.3%), with a mean age of 63.5 ± 3.1 years. The majority of patients (65.9%) were in a critical care unit; 21.9% were on a specialty floor; and 12.2% were on “other” units. Most patients had a primary diagnosis that was surgical (70.4%). Most patients were turned using a drawsheet (53.6%), and most patients (69.4%) assisted with the turn. Only 12% of the patients were premedicated for the turn with an opioid analgesic.

Table 1. Demographic and Clinical Characteristics of the Sample (n = 1,395; Mean Age 63.5 yrs, Range 18-97 yrs)

		n	%
	Gender
	Female	592	44.7%
	Male	803	55.3%
	Ethnicity
	White	1204	86.3%
	Black	98	7.1%
	Other	77	5.6%
	Level of care
	Critical care	907	65.9%
	Specialty floor	302	21.9%
	Other	167	12.2%
	Primary diagnosis
	Medical	332	24.2%
	Surgical	965	70.4%
	Trauma/burn/other	74	5.4%
	Turning method
	Hands	518	43.6%
	Drawsheet	637	53.6%
	Other	34	2.8%
	Patient assist with turn
	Assisted	927	69.4%
	Did not assist	409	30.6%
	Opioid premedication
	Yes	168	12.0%
	No	1227	88.0%

Pain Intensity Reports 

The mean pain intensity score at the time of the turn was 4.9 ± 3.1. As shown in Table 2, except for ethnicity, significant differences were found in mean pain intensity scores for a variety of demographic and clinical characteristics. Specifically, women, specialty floor patients, and surgical and trauma or burn patients reported significantly higher levels of pain during the turn. Those who were turned with a drawsheet, those who did not assist with the turn, and those who were premedicated reported more pain.

Table 2. Differences in Mean Pain Intensity Score During Turning According to Demographic and Clinical Characteristics (n = 1,395)

		Mean	(SD)	Statistics
	Overall	4.9	(3.1)
	Age (yrs)			F = 22.2∗∗∗
	18-40	6.0	(2.8)
	41-60	5.5	(3.0)
	≥61	4.5	(3.1)
	Gender			t = 3.9∗∗∗
	Male	4.7	(3.0)
	Female	5.2	(3.2)
	Ethnicity			F = .01
	Black	4.9	(3.4)
	White	4.9	(3.1)
	Other	4.9	(3.0)
	Level of care			t = 3.9∗∗∗
	Critical care	4.7	(3.1)
	Specialty floor/other	5.4	(3.0)
	Diagnosis			F = 34.6∗∗∗
	Medical	3.8	(3.4)
	Surgical	5.2	(2.8)
	Trauma/burn/other	6.1	(3.1)
	Turn procedure			t = 3.0∗∗∗
	Drawsheet	5.2	(3.1)
	Hands/other	4.7	(3.0)
	Patient assist with turn			t = 2.8∗∗∗
	Assisted	4.8	(3.0)
	Did not assist	5.3	(3.2)
	Opioid premedication			t = −6.5∗∗∗
	Premedicated	6.4	(2.8)
	Not premedicated	4.7	(3.1)

∗∗∗ p < .001.

Frequency of Nonpharmacologic Interventions 

Table 3 reports the frequency of specific nonpharmacologic interventions used during turning. The most frequently reported interventions included calming voice, information, deep breathing, gentle touch/hand holding, distraction, pillow splinting, and humor. Somewhat less frequently used were massage, presence of family/friends, therapeutic touch, progressive relaxation, ice, and “other.” Infrequently used interventions were music, heat, guided imagery, TENS, acupressure, and hypnosis.

Table 3. Frequency of Nonpharmacologic Interventions Used During Turning (n = 1,395)

		%	n
	Calming voice	65.7%	917
	Information	60.6%	845
	Deep breathing	37.9%	528
	Gentle touch/hand holding	36.6%	510
	Distraction	34.2%	474
	Pillow splinting	34.0%	473
	Humor	25.9%	361
	Massage	15.4%	215
	Presence of family/friends	13.7%	191
	Therapeutic touch	10.1%	141
	Progressive relaxation	8.1%	113
	Other	4.7%	65
	Ice	2.3%	32
	Music	1.9%	27
	Heat	1.4%	20
	Guided imagery	0.9%	13
	TENS	0.7%	10
	Acupressure	0.2%	3
	Hypnosis	0.0%	0

TENS = transcutaneous electrical nerve stimulation.

Figure 1 presents the total number of all nonpharmacologic interventions used during one particular turn. The vast majority of patients (92.5%) had at least one nonpharmacologic intervention, while over one-fourth of patients (26.3%) had five or more interventions.

View Large Image Download to PowerPoint

Figure 1 Number of nonpharmacologic interventions used during one turn (n = 1,395).

The three most frequently performed nonpharmacologic interventions were calming voice (65.7%), receiving information (60.6%), and deep breathing (37.9%). Table 4 presents the differences in use of each of these three interventions according to patients' demographics and clinical characteristics. Use of calming voice differed according to whether the patient was in a critical care unit. Patients in critical care units were more likely to receive a calming voice. Those who were turned with the use of a drawsheet and those who reported a higher pain intensity more frequently received a calming voice. Those who were in a critical care unit, who did not assist the health care provider during the turn, who had higher pain intensity, or who were nonmedical patients received information significantly more frequently. Use of deep breathing was the only nonpharmacologic intervention that was influenced by age. Those patients who used deep breathing were younger than those who did not. Deep breathing was used more frequently by those who reported a higher pain intensity score, were in a critical care unit, had a surgical versus trauma or burn diagnosis, and were premedicated with opioids.

Table 4. Differences in Demographic and Clinical Characteristics for Each of the Three Most Frequently Used Nonpharmacologic Interventions (n = 1,395)

		Calming Voice		Information		Deep Breathing
	Gender
	Male	62.9%		63.6%∗		39.2%
	Female	67.4%		55.7%		35.8%
	Ethnicity
	White	66.6%		62.4%∗		38.9%
	Black	68.8%		55.3%		30.8%
	Other	53.3%		45.2%		38.1%
	Care level
	Critical care	68.2%∗		65.2%∗∗		40.5%∗∗
	Specialty floor/other	60.6%		51.3%		32.8%
	Diagnosis
	Medical	62.5%		47.9%∗∗		18.8%∗∗
	Surgical	66.3%		64.3%		44.5%
	Trauma/burn/other	76.8%		75.0%		38.7%
	Turn method
	Hands	58.5%∗∗		56.7%∗∗		39.0%
	Drawsheet	71.2%		66.0%		38.5%
	Other	54.8%		35.5%		32.5%
	Patient assist with turn
	Assisted	64.5%		57.4%∗∗		37.0%
	Did not assist	70.1%		68.0%		41.8%
	Opioid premedication
	Premedicated	70.7%		67.3%		57.6%∗∗
	Not premedicated	65.0%		59.7%		32.5%
		Yes	No	Yes	No	Yes	No
	Mean NRS pain (during turn)	5.3	4.4∗∗	5.2	4.6∗∗	5.6	4.6∗∗
	Mean age (yrs)	63.7	63.1	63.6	63.3	61.1	65.0∗∗

NRS = numeric rating scale. Chi-square tests for categoric variables and t tests for continuous variables

∗ p < .01. ∗∗ p < .001.

Predictors of the Three Most Frequently Performed Nonpharmacologic Interventions 

Separate multivariate logistic regression models were constructed for calming voice, deep breathing, and information (Table 5). Because all demographic and clinical variables had significant univariate associations with at least one of the three top nonpharmacologic interventions, all demographic and clinical predictors were included in each model. Age did not predict use of any of the three nonpharmacologic interventions studied. Gender was predictive of use of a calm voice: Female patients were 1.3 times more likely than male patients to receive a calming voice (odds ratio [OR] 0.74, 95% confidence interval [CI] 0.56-0.98). Ethnicity was predictive, with white patients 1.7 times more likely than patients of other ethnicities to receive a calming voice (OR 1.74, 95% CI 1.01-3.02) and 1.9 times more likely than patients of other ethnicities to receive information (OR 1.87, 95% CI 1.07-3.25). Those patients in critical care were more likely to receive each of these nonpharmacologic interventions than were those on a specialty floor (calming voice: OR 1.66, 95% CI 1.23-2.23; information: OR 1.62, 95% CI 1.21-2.16; deep breathing: OR 1.36, 95% CI=1.01-1.84).

Table 5. Multivariate Logistic Regression of Demographic and Clinical Predictors for Each of the Three Most Frequently Performed Nonpharmacologic Interventions During Turning (n = 1,395)

		Calming Voice		Information		Deep Breathing
		OR	95% CI	OR	95% CI	OR	95% CI
	Age	1.00	0.99-1.01	1.01	0.99-1.01	0.99	0.98-1.00
	Male gender	0.74∗	0.56-0.98	1.23	0.93-1.62	1.01	0.76-1.34
	Ethnicity
	Black	1.68	0.81-3.50	1.27	0.61-2.64	1.21	0.55-2.66
	White	1.74∗	1.01-3.02	1.87∗	1.07-3.25	1.43	0.79-2.62
	Critical care	1.66∗∗	1.23-2.23	1.62∗∗∗	1.21-2.16	1.36∗	1.01-1.84
	Diagnosis
	Surgical	1.02	0.73-1.42	1.73∗∗∗	1.25-2.38	2.33∗∗∗	1.62-3.34
	Trauma/burn/other	1.70	0.82-3.53	4.12∗∗∗	1.97-8.63	1.56	0.78-3.11
	Turned with drawsheet	1.63∗∗∗	1.23-2.14	1.53∗∗	1.17-2.02	0.90	0.68-1.19
	Patient assisted with turn	1.01	0.74-1.37	0.86	0.64-1.17	0.85	0.64-1.16
	Pain intensity during turn	1.01∗∗∗	1.01-1.02	1.01∗	1.001-1.009	1.01∗	1.001-1.010
	Opioid premedication	1.05	0.68-1.61	1.11	0.73-1.69	1.58∗	1.06-2.36

CI = confidence interval; OR = odds ratio. Reference categories: ethnicity: “other”; level of care/critical care: “specialty floor, other”; diagnosis: “medical”; turning method: “turned with hands/other.”

∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001.

Primary diagnosis predicted the use of nonpharmacologic interventions. Surgical patients were more likely than medical patients to receive information (OR 1.73, 95% CI 1.25-2.38) and deep breathing (OR 2.33, 95% CI 1.62-3.34) interventions. Patients who reported a higher pain intensity score were more likely to receive each of these nonpharmacologic interventions (calming voice: OR 1.01, 95% CI 1.01-1.02; information: OR 1.01, 95% CI 1.001-1.009; deep breathing: OR 1.01, 95% CI 1.001-1.010). For example, this means that the odds of using a calming voice was 1.01 times more likely for each 1-point increase in the 10-point NRS.

Discussion 

The results of this study provide important new insights into the use of nonpharmacologic interventions in the hospital setting for adult patients who are being turned. Of note, nonpharmacologic interventions were used more frequently than opioid premedication. The nonpharmacologic interventions most frequently used were calming voice, receiving information, and deep breathing. In addition, for these interventions, clinical characteristics (such as the patient's diagnosis and level of care) as well as some demographic characteristics were associated with the use of nonpharmacologic interventions.

Most notably, nonpharmacologic interventions were used frequently, with 93% of patients receiving at least one and 26% receiving five or more. Although no data were found in the literature on frequency of nonpharmacologic interventions for procedural pain, the limited data that exist on their use for acute pain suggest that these techniques are not used frequently (Carroll et al., 1999, Manias et al., 2005). However, the Thunder II study differed from those studies by including a very comprehensive list of nonpharmacologic interventions, which might partially explain this discrepancy. In addition, nonpharmacologic interventions may be used more frequently for procedural pain than for acute pain. One reason for their increased use with procedures may be that a nurse is present for the procedure and is likely to initiate or coach the patient to use a nonpharmacologic intervention, such as deep breathing, during procedures. Another possible explanation is that procedures, especially turning, occur during brief, discrete, and relatively predictable blocks of time, which could facilitate their use.

Of concern is our finding that few patients were premedicated with analgesics. Because turning usually requires a second person, making the exact timing unpredictable, nurses may sometimes have insufficient opportunity to prepare for premedication. Therefore, when patients experienced pain during turning, they and their caregivers may have attempted to control the pain using other methods. The nonuse of medications is clearly problematic, given the amount of pain experienced by these patients who were turned. Nursing vigilance and attention to the importance of premedication whenever possible may help to decrease pain associated with this ubiquitous procedure. Furthermore, the combined use of pharmacologic and nonpharmacologic interventions during a turning procedure may provide the optimal approach to pain management and warrants further study.

The most frequently used interventions during turning included calming voice, information, deep breathing, gentle touch/hand holding, distraction, pillow splinting, and humor. In general, these techniques are easy to implement and did not require equipment or specific training. In contrast, the methods that require equipment and/or training, such as TENS, music, and guided imagery were not used frequently. These results are not unexpected, given the busy nature of acute and critical care settings.

Overall, the nonpharmacologic interventions previously studied for efficacy do not fully correspond to those that were most frequently used in the present study. For example, the largest body of research is on the use of music for procedural pain (Bally et al., 2003, Broscious, 1999, Cepeda et al., 2006, Chan et al., 2006, Kwekkeboom, 2003), but only 1.9% of the present sample used music. Calming voice, gentle touch, pillow splinting, and humor were four of the six most frequently used interventions, yet no data were found regarding their effectiveness for procedural pain. Although this study did not examine the effectiveness of these interventions for pain, the identification of and predictors of their use provides foundational information for future studies. Of the most frequently used interventions in our study, only information and deep breathing were evaluated in earlier studies.

Lang et al. (2005) examined the relationship between pain intensity and information about painful sensations, such as “sharp” or “stinging,” patients might feel during interventional radiologic procedures. Interestingly, patients who were told to expect pain or noxious stimuli reported higher procedural pain intensity scores than those patients who were not told what to expect. In the second study, Puntillo et al. (2004) examined the addition of sensory information to a pharmacologic analgesic intervention for chest tube removal. Sensory information was that which described the specific sensation to be expected, such as aching, sharp, or stinging. No differences were found in pain intensity and distress between patients who received sensory information with an analgesic and those who received only an analgesic. Findings from these two studies suggest that information may not alleviate procedural pain.

Two studies assessed the effectiveness of deep breathing on procedural pain. In the first, Friesner et al. (2006) found that patients who performed deep breathing exercises (i.e., inhaling through nose and exhaling slowly through pursed lips) and received opioids had a significant decrease in pain intensity scores during chest tube removal compared with patients who received only opioids. In a second randomized clinical trial (Lang et al., 2000), patients undergoing percutaneous vascular and renal procedures randomized to a self-hypnosis group (i.e., instructed to close their eyes, breath deeply, and concentrate on a sensation of floating; n = 82) reported a constant level of procedural pain. This was in comparison to increases in pain in patients in attention control (n = 80) or standard care groups (n = 79). This finding suggests that a hypnosis intervention which includes deep breathing may help to diminish this type of procedural pain. Because the hypnotic intervention included eye closing and concentration on a sensation of floating, it is not possible to determine the independent contribution of deep breathing. Yet, both of these studies provide some evidence that deep breathing may help to alleviate procedural pain.

Regarding the efficacy of the most frequently used nonpharmacologic interventions in our study, the evidence is mixed. No data are available on the effectiveness of a calming voice, information may have no effect or a detrimental effect on pain, and deep breathing may help to alleviate pain. All of these data focus on procedures other than turning. Given the paucity of research, it is unclear that the information obtained about one procedure is generalizable to another. Randomized controlled trials are needed to elucidate the efficacy of different nonpharmacologic interventions, with subsequent development of evidence-based guidelines.

The present study explored demographic and clinical characteristics that, when studied within the context of other factors, influenced whether or not a patient got the interventions of calming voice, deep breathing, and information. The logistic regression analyses indicated that in general, clinical factors did affect nonpharmacologic intervention use, and only some demographic factors did. Of demographic variables, age was not predictive. This suggests that, although age is inversely related to level of reported pain, age was not associated with the use of nonpharmacologic interventions for turning. Of the clinical characteristics, a critical care setting and pain intensity during turn were consistently associated with increased likelihood that a patient would receive each of the three nonpharmacologic interventions. Perhaps the finding related to setting can be explained by increased nursing experience and better nurse-patient ratios in ICUs. A critical-care nurse is often more experienced and cares for fewer patients. Therefore, the critical-care nurse may be able to allocate more time for the turn and be more comfortable and skilled in using multiple interventions. The level of a patient's pain intensity may be related to nonpharmacologic intervention use; nurses who know that a patient is having pain during the procedure may be aware of a patient's pain during turning and respond with nonpharmacologic interventions that are immediately available. Similarly, when patients experience procedural pain during a turn, they themselves might initiate an intervention, such as deep breathing.

Although this study presents new findings, it has limitations. Because data were collapsed across patients', nurses', and others' responses, we do not know for sure who initiated the nonpharmacologic interventions and whether one provider is more successful that another in using these interventions. Yet, we thought that identity of performer was less important than capturing the overall use of nonpharmacologic interventions in this exploratory study. Owing to the observational design of the study, we could not evaluate the effectiveness of the various nonpharmacologic interventions that were used for turning. The finding that pain was positively associated with use of a nonpharmacologic intervention underscored this limitation. Another limitation is that instead of random sampling, a convenience sample was recruited. This could contribute to selection bias and limit the generalizability of the findings. However, the sample was very large, with multiple units in many hospitals.

Although the large sample size based on multiple sites strengthens generalizability, it can lead to statistical overpowering and the finding of significant results that are due to chance. However, some of the key findings, such as frequency of use of nonpharmacologic interventions, were purely descriptive and not influenced by statistical power. For the univariate and logistic regression analyses, most significant findings met a more rigorous level of significance (p < .01 or p < .001). Another potential limitation of the sample is that it was not very ethnically diverse: it was predominantly white (86.3%), with a small group of black patients (7.1%). This ethnic distribution limits applicability of findings to Asian, Latino, or other ethnically diverse patient populations. Future studies should incorporate sampling interventions to ensure a more ethnically diverse sample.

Conclusions 

This study found that nonpharmacologic interventions were used frequently for turning. The specific interventions used most often included calming voice, information, and deep breathing, ones that can be initiated spontaneously and without specific equipment or training. These data suggest that patients, nurses, and family members may be aware of patients' pain during turning and respond to their increased pain by using nonpharmacologic interventions available in that situation, such as calm voice and deep breathing, Randomized controlled trials are needed that examine the effectiveness of nonpharmacologic interventions for procedural pain, especially considering that some are frequently used in clinical practice.