To the Editor:—
Pain is one of the major stressors experienced by the patients hospitalized in an intensive care unit (ICU).1The DOLOREA study has shown that 33% of the ventilated patients experienced pain at rest and 56% experienced pain during a procedure.2Recently, we have reported that a systematic evaluation of pain and agitation at rest in ventilated and nonventilated ICU patients was associated with a better outcome.3Although literature evaluating procedural pain is consistent,4surprisingly, there are few data available regarding the occurrence of pain at rest in ICU patients.5The objective of the current analysis was to compare the incidence and characteristics of pain at rest in surgical and trauma versus medical ICU patients included in the previous database.3
All consecutive patients aged 18 yr or older and staying in a 12-bed medical–surgical ICU for more than 24 h were eligible. Exclusion criteria were decision to withdraw life support within 48 h after admission, brain injuries that limited communication by the patient, and transfer to another ICU for specialized care. Pain and agitation scores were recorded twice daily by nurses or students in medicine or pharmacy at rest, 30 min after any procedure. The 0- to 10-point numerical rating scale,6enlarged to be easily visible (3.9 × 11.8 inches), was used. The behavioral pain scale score7,8was used for evaluation of pain in intubated or tracheotomized patients if they were not able to perform the numerical rating scale. Only moderate to severe pain events were recorded. Therefore, a pain event was defined by either a behavioral pain scale score greater than 5 according to the study of Payen et al .7or a numerical rating scale score greater than 3 according to usual definitions.6The main cause of pain was prospectively documented in communicating patients. Vigilance and agitation had been assessed with the French-translated Richmond Agitation–Sedation Scale.9Data were prospectively recorded as previously described.3Trauma and surgical patients were grouped together (group ST) and compared with the medical patients (group M). Quantitative data are shown as median [25th–75th percentiles]. Univariate analyses (chi-square, Fisher test, Mann–Whitney U test) between the two groups were used. A P value of 0.05 or less was considered statistically significant.
A total of 230 patients were included for analysis, 154 in group ST (12 trauma, 142 postoperative patients) and 76 in group M. Among the 142 postoperative patients, 77 were admitted to the ICU after an unplanned surgery, 47 were admitted after a planned surgery, and 18 were admitted after a postoperative complication that occurred at a median time of 3.5 [3.0–5.0] days after surgery. The abdominal site was the site of surgery for 136 of the 142 postoperative patients. Reasons for admission for medical patients were acute respiratory failure (n = 24), drug intoxication (n = 12), digestive bleeding (n = 9), acute pancreatitis (n = 8), septic shock (n = 8), acute renal failure (n = 5), and miscellaneous (n = 10).
The incidence of pain in the 230 evaluated patients was 51%, with no significant difference between group ST and group M (52% vs . 50%; P = 0.78). The number of pain ratings was not significantly different between the two groups (11.0 [5.3–19.0]vs . 9.0 [5.0–18.3]; P = 0.43). Group ST had a significantly higher rate of intubation (77% vs . 55%; P < 0.001), a lower Simplified Acute Physiology Score II (29 [20–39]vs . 36 [26–49]; P < 0.01), and a lower sepsis rate (36% vs . 49%; P = 0.05) at admission. The use of analgesic drugs before the diagnosis of pain was significantly greater in group ST (64% vs . 37%; P = 0.0001). Acetaminophen was the main drug used in this situation (86% in the two groups). Fourteen surgical patients had epidural analgesia. No significative difference was shown between group ST and group M for age (58 [50–70]vs . 58 [47–73] yr), female sex (33% vs . 34%), duration of mechanical ventilation (96 [24–192]vs . 132 [36–288] h), use of a continuous infusion of sedatives (57% vs . 50%), its duration (54 [24–144]vs . 96 [24–204] h), Richmond Agitation–Sedation Scale level of sedation (−4.1 [−2.5 to −4.8]vs . −4.2 [−3.6 to −4.3]), duration of stay (8.0 [4.0–13.5]vs . 6.0 [3.0–13.7] days), or mortality (12% vs . 17%) in the ICU.
Table 1shows the characteristics of pain in all patients with pain. No significative difference was shown between the two groups except for the median intensity of the numerical rating scale score, which was significantly higher in group M than in group ST (5.6 [5.0–6.7]vs . 5.0 [4.3–6.0]; P = 0.03). Figure 1reports the causes of pain in the communicating surgical-trauma (n = 71) and medical (n = 34) patients. The site of injury responsible for admission is the main cause of pain at rest (49%) for surgical-trauma patients, whereas the back and limbs were the main causes of pain at rest (41%) in medical patients.
The main finding of this analysis is that the incidence of pain in this population of ICU medical patients is not different than that in surgical-trauma patients. Moreover, intensity of pain in ICU medical patients experiencing pain is significantly higher than for surgical-trauma patients. Medical patients received preventive analgesia less frequently. The back and limbs are the main areas of pain in medical patients. This could be explained in part by the fact that medical patients had a greater rate of sepsis upon admission than the surgical patients. Myalgia and arthralgia are common clinical features associated with fever, determined in part by inflammation and the muscle hypercatabolism induced by the thermogenesis.10Inflammatory cytokines and sympathetic amines have been implicated in the hypernociceptive state associated with inflammation.11Back and limb pain may be also related to the obligatory immobilization of patients in the ICU bed, often requiring the use of sedatives or physical restraint.12Considering that turning of the patient is the most painful procedure in the ICU2,4and that the pain before the procedure is associated with pain during the procedure,5,13efforts to better manage pain at rest should be encouraged. Moreover, decreasing pain at rest and its stress response may be associated with a better outcome in critically ill patients.3If so, pain at rest should be considered as a major clinical diagnostic symptom.
In conclusion, the incidence and intensity of pain in ICU medical patients are not lower than in surgical-trauma patients in this cohort of ventilated and nonventilated patients.
*Saint Eloi Hospital, Montpellier University Hospital, Montpellier, France. email@example.com