Article Text

other Versions

PDF
Patient-controlled analgesia compared with interval analgesic dosing for reducing complications in blunt thoracic trauma: a retrospective cohort study
  1. Stephen Edward Asha1,2,
  2. Kate Anne Curtis3,4,5,
  3. Colman Taylor5,
  4. Allan Kwok2,6
  1. 1Department of Emergency, St George Hospital, Sydney, New South Wales, Australia
  2. 2Faculty of Medicine, University of New South Wales, Sydney, New South Wales, Australia
  3. 3Trauma Service, St George Hospital, Sydney, New South Wales, Australia
  4. 4Sydney Nursing School, University of Sydney, Sydney, New South Wales, Australia
  5. 5The George Institute for Global Health, Sydney, New South Wales, Australia
  6. 6Department of Surgery, St George Hospital, Sydney, New South Wales, Australia
  1. Correspondence to Dr Stephen Edward Asha, Department of Emergency, St George Hospital, Gray St, Kogarah, NSW 2217, Australia; stephen.asha{at}sesiahs.health.nsw.gov.au

Abstract

Objectives To determine if complications from blunt thoracic trauma are reduced with patient-controlled analgesia (PCA) compared with interval analgesic dosing given as needed. Secondary aims were to investigate the influence of PCA on hospital length of stay (LOS) and cost.

Methods In this retrospective cohort study, patients were identified using the hospital trauma registry and clinical information department. Data on analgesic method, outcomes and confounders were obtained from the medical record. Costing data were obtained from the case-mix department. The analysis used logistic regression for the primary outcome and a generalised linear model for the secondary outcomes to adjust for potential confounders.

Results 227 patients were included. In the PCA group, 17/52 (33%) patients had a complication compared with 26/175 (15%) in the interval dosing group. The adjusted odds for a complication in patients receiving PCA was not significantly different from the adjusted odds in those receiving interval dosing (OR=1.2, 95% CI 0.3 to 4.6, p=0.83). The median LOS was 8.9 days in the PCA group and 4.6 days in the interval dosing group. The adjusted LOS for patients receiving PCA was 10% shorter than those receiving interval dosing (relative difference 0.9, 95% CI 0.6 to 1.3, p=0.52). The median hospital cost was $A11 107 in the PCA group (IQR $A7520–$A15 744) and $A4511 (IQR $A2687–$A8248) in the interval dosing group. The adjusted total hospital costs for patients receiving PCA was 10% higher than for those receiving interval dosing (relative difference 1.1, 95% CI 0.8 to 1.5, p=0.44).

Conclusions PCA did not reduce complications, hospital LOS or costs compared with interval analgesic dosing.

Statistics from Altmetric.com

Introduction

Blunt chest trauma is a common presentation to emergency departments (EDs) and one of the most common injuries amongst patients with trauma. It can be associated with a high death rate in the older population1 regardless of injury severity and comorbidity.2–4 Patients with isolated chest trauma and at least three rib fractures have an increased risk of in-hospital death than those with blunt chest trauma and fewer than three rib fractures. This effect is more dramatic for those with more than four rib fractures5 and older patients.6 The pain associated with thoracic trauma results in poor mobility and poor lung expansion and this can lead to complications such as venous thromboembolism, pneumonia and respiratory failure, with the need for respiratory support. Effective pain control to allow deep breathing, chest physiotherapy and improved lung function7 ,8 are essential aspects of thoracic trauma care and the prevention of complications.9 ,10

Patient-controlled analgesia (PCA) has been shown to provide more effective pain relief than interval analgesic dosing provided as needed when requested by the patient or initiated by nursing staff11 among patients with postoperative pain. Extrapolating from this information, it is probable that PCA would also provide more effective pain relief than interval analgesic dosing in patients with blunt thoracic trauma. However, it is not known if the more effective analgesia provided by PCA leads to fewer complications from thoracic trauma and a reduction in healthcare resource use. PCA use is common on wards after admission but its use in the ED is infrequent, possibly owing to the perceived time constraints for emergency nursing staff in the set up of equipment. A stronger case could be made for the early use of PCA while patients were still in the ED if there was evidence that it provided both an analgesic and a therapeutic benefit.

We hypothesise that the use of PCA in patients with blunt thoracic trauma reduces complications in comparison with patients who receive interval analgesic dosing given as needed. The primary aim of this analysis was to investigate the influence of PCA on complication rate. Secondary aims were to investigate the influence of PCA on hospital length of stay (LOS) and cost.

Methods

This retrospective cohort study was undertaken at St George Hospital, a level 1 trauma centre located in Sydney, New South Wales, Australia, between October 2008 and February 2010. Permission for the study was granted by the hospital's human research ethics committee. St George Hospital admits more than 50 000 patients annually and it is the trauma referral hospital for south eastern Sydney, a geographic zone of about 1.5 million inhabitants. In 2009, 1700 patients fulfilled trauma team activation criteria. Of these, about 1000 required admission and over 250 of these were severely injured with an Injury Severity Score (ISS) >15.

Patients were eligible for the study if they were admitted to the hospital with isolated blunt thoracic trauma. The inclusion and exclusion criteria are outlined in box 1. These criteria were chosen to limit participants to those who essentially had isolated thoracic trauma and would have been suitable to use PCA. Patients with limb trauma requiring an operation were excluded as they might have received PCA for analgesia of the limb rather than for thoracic trauma.

Box 1

Eligibility criteria

Inclusion criteria:

  • Blunt thoracic trauma mechanism

  • Adult ≥18 years

  • Admitted to hospital

  • Rib and/or sternum fracture (clinical or radiological diagnosis)

Exclusion criteria:

  • Major head, neck, abdomen or pelvis trauma (AIS >2 for any of these regions, or injury of these regions requiring operative intervention)

  • Limb trauma requiring operative intervention

  • Intubated in the emergency department

  • Spinal or epidural analgesia or intercostal block

  • AIS, Abbreviated Injury Score.

Potential participants were identified by searching the hospital clinical information database using ICD 10 codes related to thoracic trauma. The database of the hospital trauma service was also searched for additional patients using the abbreviated injury scale.12 The medical records were then retrieved and screened for inclusion and exclusion criteria. Those patients who fulfilled the eligibility criteria underwent a structured chart review performed by three of the authors (SEA, KAC, AK). A standardised template for data extraction was created and refined through several trials on a small sample of medical records. Each data point was clearly defined and the section of the medical record from which the data item was to be obtained specified. Radiographic data were obtained by retrieving the reports from the radiology database. Hospital costing data were obtained from the hospital case-mix unit. Patient costing, including indirect expenses (ie, overheads, human resources using staffing head count, cleaning expense using floor space) was conducted in accordance with 2008–9 NSW program and product data collection.13

The cohort was divided into two groups—those who were allowed PCA and those who received interval analgesic dosing—and comparisons were made for the outcomes of interest with adjustment for the effects of potential confounders. A patient was placed in the PCA group if they had received PCA at any time during their hospital stay. The primary outcome was a composite of all-cause mortality, pneumonia, deep vein thrombosis, pulmonary embolism, need for ventilatory support after admission (continuous positive airway pressure, bilevel positive airway pressure or endotracheal intubation) and unplanned admission to the intensive care unit. The secondary outcomes were hospital LOS and total cost of the hospital stay. Pneumonia was determined to have occurred if there was radiological evidence of pulmonary air-space opacification together with documentation in the medical record stating a diagnosis of pneumonia and treatment with antibiotics. When radiological evidence of pulmonary air-space opacification developed within 24 h of arrival, these changes were considered to represent contusion rather than infection. Potential confounders evaluated were age, sex, comorbid disease (cardiac, respiratory, neurological, renal and liver disease, diabetes, osteoporosis and active cancer), anticoagulant and antiplatelet drug use, mechanism of injury, physiotherapy review, pain service review, service under which admitted, thoracic imaging performed, associated thoracic injuries (pulmonary contusion, pneumothorax, haemothorax), thoracostomy tube insertion, ISS, new ISS, vital signs and number of rib fractures. Potential confounders were selected by clinical opinion and previous research.3 ,14 ,15

The statistical analysis plan was determined a priori. Data were analysed with SAS statistical software V.9.2 (SAS Institute, Cary, North Carolina, USA). To examine the influence of PCA on the three outcomes, three different statistical models were used. The influence of PCA on complication rate was examined using logistic regression. The influence of PCA on hospital LOS was examined using a generalised linear model with a log link, whereas the influence of PCA hospital cost was examined using a generalised linear model with a log link and a gamma distribution to account for the non-negativity and positively skewed distribution of costing data. In each model potential confounders were examined independently and those with a p value <0.2 in a univariate analysis were included in the multivariate model, with the exception of age which was included regardless of the p value.

Results

Of the 227 patients who satisfied the inclusion and exclusion criteria, 52 (23%) received PCA and 175 (77%) received interval analgesic dosing. Patients in the PCA group received PCA treatment for a median of 2.6 days (IQR 1.3–4.4). The minimum time on PCA was 0.5 days and the maximum was 9.0 days. Expressed as a proportion of total hospital stay, the median proportion of time patients in the PCA group received PCA was 38% (IQR 19–57%), with a minimum of 4% and a maximum of 95%. Patient who received PCA tended to be younger, have more severe injuries (higher ISS) and have higher-energy injury mechanisms than those receiving interval dosing analgesia. These baseline characteristics are presented in tables 1 and 2, which outline the univariate associations between potential confounding variables and each of the primary and secondary outcome variables. Smoking status was not evaluated owing to a large proportion of missing values.

Table 1

Characteristics of participants according to PCA analgesic status.

Table 2

Univariate analysis of the association with the outcome variables*

Seventeen patients (33%, 95% CI 20% to 47%) in the PCA group had a complication, compared with 26 (15%, 95% CI 10% to 21%) in the interval dosing group. After adjusting for confounding variables, the odds of a complication occurring in patients receiving PCA was 1.2 (95% CI 0.3 to 4.6, p=0.83) times higher than the odds of a complication occurring among those receiving interval dosing. The median hospital LOS was 8.9 days (IQR 5.4–11.7) in the PCA group and 4.6 days (IQR 2.0–8.0) for the interval dosing group. After adjusting for the effects of confounding variables, the hospital LOS for patients receiving PCA was 10% shorter than the LOS of patients receiving interval dosing (relative difference 0.9, 95% CI 0.6 to 1.3, p=0.52). The median cost of the hospital stay was $A11 107 in the PCA group (IQR $A7520–$A15 744) and $A4511 (IQR $A2687–$A8248) in the interval dosing group. After adjusting for the effects of confounding variables, the total hospital costs for patients receiving PCA was 10% higher than those for patients receiving interval dosing (relative difference 1.1, 95% CI 0.8 to 1.5, p=0.44).

Given the null result we undertook a further exploratory analysis of the primary outcome, examining the influence of PCA in a more severely injured patient population defined by more than two rib fractures. One hundred and eight patients had more than two rib fractures—39 (36%, 95% CI 27% to 46%) received a PCA and 69 (64%, 95% CI 54% to 73%) received interval analgesic dosing. In this more severely injured subgroup, after adjusting for confounding variables, the odds of a complication occurring in patients receiving PCA was 1.6 (95% CI 0.5 to 5.8, p=0.46) times higher than the odds of a complication occurring among those receiving interval dosing.

Discussion

In this study we were unable to demonstrate a benefit from the use of PCA in patients with blunt thoracic trauma, compared with patients receiving interval analgesic dosing as needed when requested by the patient or initiated by nursing staff. While the point estimates of these outcome measures tended to be in favour of interval analgesic dosing, these were all statistically non-significant with wide CIs and should not be interpreted as suggesting that PCA is harmful. These results probably reflect the difficulty in measuring and controlling for the many confounding effects that occur with differences in age, coexisting disease severity, injury mechanism and other variables that could not be measured and the selection bias that patients who are likely to receive PCA were younger and more severely injured. This study does not, however, provide support for our hypothesis that the more effective analgesia provided by PCA would result in fewer complications through promoting earlier mobilisation and more effective deep breathing and ventilation.

There are several other limitations to this study. Data were collected retrospectively from the medical record, and so depend on the accuracy of the information recorded. Measurement bias of the number of rib fractures might have occurred as those in the PCA group were more likely to have had chest CT because they were more severely injured and CT is more accurate in identifying rib fractures than a chest x-ray examination.16 Patients for the study were identified using hospital databases which rely on the accuracy of the coding of patient's illness/injuries. Some patients with isolated thoracic trauma night have been missed. It would also have been of value to include measures of analgesic effectiveness, such as a pain score from a visual analogue scale. Unfortunately, these data could not be obtained retrospectively from a review of the charts. Finally, the primary outcome was a composite of several complications. This composite outcome was dominated by pneumonia with other complications being relatively uncommon. The diagnosis of pneumonia can be subjective, particularly when this is being determined retrospectively. As a consequence there is likely to be some misclassification of the primary outcome measure.

The use of PCA has benefits that extend beyond the reported clinical benefits. These include patient empowerment and the reduction of nursing workload (frequently having to double check and draw up individually prescribed scheduled drugs), particularly in the acute phase of injury management in the ED.

Conclusion

In this retrospective study comparing PCA with interval analgesic dosing in blunt thoracic trauma we were unable to demonstrate any benefit from reduced complications, reduced hospital LOS or reduced costs compared with analgesia provided through interval dosing as needed. For future research we would recommend a randomised trial to further evaluate this method of analgesia for thoracic injury.

View Abstract

Footnotes

  • Contributors KAC participated in the design of the study, data collection, literature review and reviewing of the manuscript. CT performed the statistical analysis and reviewed the manuscript. AK participated in data collection and manuscript review.

  • Funding Funding for the statistical analysis was provided by the Honda automotive company via the St George and Honda trauma and critical care research programme. Honda had no role in the study conception, study design, data collection, analysis and interpretation, in the writing of the manuscript, or in the decision to submit the manuscript for publication.

  • Competing interests None.

  • Ethics approval South Eastern Sydney and Illawarra Area Health Service human research ethics committee central.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

References