Sickle cell disease (SCD) is a genetic disorder affecting >100,000 Americans and 3 million people worldwide.1 SCD accounts for over 83,000 hospital admissions each year in the United States, with longer than average inpatient hospital stays2 and frequent readmissions,3,4 at an annual hospitalization cost of $488 million and total cost of care >2 billion dollars per year.5,6 Because the majority of admissions are for pain control in the setting of acute vaso-occlusive crisis (VOC),3,7 improved methods for achieving rapid pain control are needed. Patient-controlled analgesia (PCA), in which patients can self-administer as-needed boluses of parenteral pain medications, has been associated with improved pain control when compared with standard nurse-administered intermittent bolus therapy in the postoperative setting.8 However, it is unknown if the use of PCA to treat patients admitted to medicine service with acute VOC is associated with reduced length of stay (LOS) and readmission rates. It is also unknown if the adoption of PCA would be met with mistrust by patients and lead to an increase in discharges against medical advice (AMA). Our aim was to study the effect of PCA use on patients with SCD admitted with acute VOC on inpatient LOS, and the risk of 30-day readmission and discharges AMA.
The majority of the cost associated with the care of patients with SCD is related to inpatient hospitalizations for VOC.3,7,9 In many cases, patients with SCD require multiple hospitalizations per year to control the pain associated with VOC. A retrospective multistate population-based study of >21,000 patients with SCD, reported an average LOS of 6 days and a 30-day readmission rate of 31%.2 A smaller prospective study found an average LOS of 7.6 days and it was noted that both readmission rates and LOS were related to adequacy of pain control on index admission.9 It has been postulated that improved methods of achieving pain control may decrease LOS and readmission rates.9
A key element of the treatment of acute VOC is pain control, and the typical treatment modality is parenteral opioids. Despite the paucity of data and evidence-based guidelines,1,10 the benefit of opiate therapy for the management of acute VOC is well established and recommended by medical societies.11,12 Even though there is no randomized controlled trial comparing the use of different opiate preparations in the management of VOC, clinical data suggest that pain relief can be achieved with various synthetic opioid preparations as long as the dose is adjusted to patient specific requirements.12,13 The traditional method of opioid delivery is via intermittent parenteral therapy, administered in the inpatient setting by a nurse or physician. However, in some treatment settings, PCA has replaced intermittent nurse-administered bolus therapy. Typically, PCA is delivered by a locked pump and includes the option of the patient initiating intermittent bolus therapy in prespecified amounts and intervals. The 2015 Cochrane Review of postoperative pain management found that there were lower pain intensity and improved patient satisfaction with the use of PCA, compared with conventional therapy; however, there was no improvement in hospital LOS.8
There are very few studies that address the efficacy of PCA in patients with SCD. A small study of patients with SCD in the emergency department setting compared intravenous intermittent opioids to PCA. This study did not find a change in pain control with PCA but did demonstrate shorter time to pain relief and shorter emergency department LOS among patients treated with PCA.14 Another small study, which compared intermittent intravenous opioids to PCA, did not find a difference in pain scores but did demonstrate a decrease in overall opioid usage and reduced risk of side effects among patients treated with PCA.15 Another study examined the use of PCA as part of a multifaceted bundle to improve the care of patients admitted with acute VOC. The intervention was associated with decreased inpatient LOS and no change in readmission rates, but it is not clear which elements of the bundle were responsible for the improvement.16 Our study is the largest study conducted in an inpatient setting and utilizes the most commonly used measures of health care utilization and quality of care such as LOS and 30-day readmission rate. It provides relevant and unique information that can guide pain management decisions in the future for SCD patients admitted with VOC.
Research shows that discharges AMA account for about 1% to 2% of discharges in the United States.17 It is also known that discharge AMA is associated with increased risk of readmission and short-term mortality.18,19 This type of discharge is associated with certain patient factors that include male sex, younger age, history of prior discharge AMA, lack of insurance, shorter LOS, and human immunodeficiency virus infection. The most common reason cited for discharge AMA in about a third of cases is dissatisfaction with care.20 We have used rates of discharge AMA as a surrogate marker for patient satisfaction with their care.
We sought to determine if the use of PCA is associated with a shorter inpatient LOS and decreased risk of readmission in adult patients treated for acute VOC. We hypothesized that pain treatment using PCA would be associated with shorter LOS without increased risk for 30-day readmission or discharge AMA.
Study Setting and Study Population
Montefiore Medical Center is an urban academic medical center in the Bronx, NY, affiliated with Albert Einstein College of Medicine. The medical center consists of multiple sites including Weiler (381 beds) and Moses Hospitals (706 beds). We examined clinical and administrative data on all adult admissions to the medical service at either hospital over a 3-year period from January 1, 2010, through December 31, 2012. Because some patients were admitted more than 1 time during the study period we use the term admission rather than the patient to describe the unit of analysis. Inclusion criteria were admissions to the medical service (both teaching and nonteaching services) with (1) a primary diagnosis of SCD related painful VOC and not merely for SCD associated anemia, including sickle-thalassemia syndromes and hemoglobin SC disease defined by International Classification of Diseases, Ninth Revision (ICD-9) codes that were assigned at discharge, and (2) had at least 1 order for parenteral opioids based on electronic order logs. Parenteral opioids used in this study included morphine, hydromorphone, and fentanyl administered in the form of PCA infusion, intramuscular, and intravenous boluses. Specific types and forms of analgesia were ultimately a prescriber’s choice and were influenced by previously utilized therapeutic regiments if effective and on patient’s preference.
We conducted a retrospective observational cohort study to examine the associations between use of PCA and LOS, 30-day readmission and discharge AMA, and 30-day readmission rates were used as a measure of health care utilization and the AMA rates were used as a surrogate marker of patient satisfaction with treatment. The study sample was divided into 2 groups based on the use of PCA versus standard intermittent nurse-administered bolus therapy as the modality for pain control during hospitalization.
Clinical data for each admission was extracted from a replicate of Montefiore’s Clinical Information System using Clinical Looking Glass, a quality improvement health care surveillance software.21 Montefiore’s Institutional Review Board approved the study.
The primary outcome measure was inpatient LOS, measured in hours and minutes from admission to discharge and reported in day units rounded to the second decimal point. Thirty-day readmission was defined as any admission for any reason, within 30 days of discharge, to either hospital and analyzed as both a dichotomous and time to event variable. Discharge AMA was treated as a dichotomous variable.
PCA Versus Standard Therapy
All adult admissions with a diagnosis of SCD with VOC who were treated with any form of parenteral opioid therapy were included in the study. Of those, admissions who, at any time during the admission, had an order for PCA were considered as “PCA treated” regardless of what other therapy they received. Admissions who did not receive and order for PCA were considered to have standard intermittent nurse-administered bolus parenteral therapy.
To determine if admissions treated versus not treated with PCA were similar, we examined admission demographic and clinical characteristics. Demographic characteristics of admissions included age, sex, race/ethnicity, and insurance type, analyzed as continuous or dichotomous variables. To measure the overall comorbidity of each admission we used the Charlson Comorbidity Score22,23 that was calculated using ICD-9 diagnosis codes and was analyzed as a continuous variable. We used All Patient Refined—Diagnosis Related Groups-Weight, assigned at the time of discharge to measure the overall severity of illness.
First, the 2 study groups (treated vs. not treated with PCA) were compared with respect to demographic characteristics, comorbidities, and severity of illness, using t tests, χ2, and rank-sum tests, as appropriate. Next, the median LOS of the 2 study groups was compared using a rank-sum test, because the LOS measure is not normally distributed. Thirty-day readmissions and discharges AMA were compared between the 2 study groups using χ2 tests and univariate generalized estimating equations. To determine the independent effect of PCA versus standard therapy on LOS, readmissions, and discharge AMA, generalized estimating equations were used to adjust for demographic characteristics, comorbidities, and severity of illness, and account for clustering of >1 admission per patient. In addition, time to 30-day readmission was examined using the method of Kaplan and Meier, with study groups being compared using a log-rank test.
STATA/IC software, version 13.1 (StataCorp., College Station, TX) was used for all statistical analysis and data management.
Over the study period, 823 adult admissions were admitted with a diagnosis of SCD with VOC and were treated with parenteral opioid therapy. Of the sample, 536 (65.1%) admissions were treated using PCA, and 287 (34.9%) were treated using standard intermittent nurse-administered parenteral therapy.
Admissions treated using PCA were similar to admissions treated with standard intermittent nurse-administered parenteral therapy with respect to age, sex, race/ethnicity, and clinical characteristics (Table 1).
Treatment with PCA was associated with significantly shorter median LOS, compared with admissions treated with standard therapy (7.46 vs. 9.42 d, P=0.001). This represented a significant difference of 1.74 days shorter LOS in admissions treated with PCA (P=0.02). Treatment with PCA remained associated with shorter LOS after adjustment for demographic and clinical characteristics, but the difference was no longer significant (difference: 1.47 d, P=0.06) (Table 2).
30-Day Readmission Rate
The 30-day readmission rate was significantly lower for the group of admissions that utilized PCA versus standard therapy (36.7% vs. 47%, P=0.004). This represented a significant reduction in risk in the unadjusted model (odds ratio [OR]unadj: 0.69; 95% confidence interval [CI]: 0.54-0.89, P=0.004), but the difference was not significant (ORadj: 0.76; 95% CI: 0.54-1.06, P=0.11) after adjustment for demographic and clinical characteristics (Table 3).
Admissions treated with PCA versus standard therapy had similar rates of discharge AMA (13.1% vs. 11.5%, P=0.52). Odds of discharge AMA when treated with PCA were similar in unadjusted (ORunadj: 1.10; 95% CI: 0.69-1.76, P=0.68), and adjusted analysis (ORadj: 1.19; 95% CI: 0.73-1.94, P=0.49) (Table 3).
In this cohort study, we found that, compared with standard intermittent nurse-administered parenteral therapy, PCA was associated with reduced LOS among adult patients with SCD admitted with acute VOC. However, the reduction in LOS (1.47 d) was no longer statistically significant (P=0.06) in a model that adjusted for demographic and clinical characteristics and accounted for clustering. Treatment with a PCA was not associated with increased risk for 30-day readmission, and discharge AMA. Although our findings are of borderline significance, we found that treatment with PCA may be associated with shorter LOS and decreased 30-day readmission rate. This is one of few studies to demonstrate the effectiveness of PCA use on inpatient floors and we believe that treatment with PCA, in addition to other non-opioid–based therapies, should be considered a possible first-line therapy for adult patients admitted with SCD and acute VOC.
Reduction in the LOS for patients with SCD admitted with VOC has been the goal of several published research studies and hospital-based quality improvement initiatives. In a study by Gonzalez et al,14 PCA use was associated with decreased LOS in the emergency department. Another study by van Beers et al,15 found that PCA use was associated with shorter LOS among patients admitted with VOC, but the difference was not significant. Shorter LOS is desirable because it may reflect timely pain control,8 may expose patients to a decreased risk of complications associated with hospitalization,24 and may free inpatient beds for patients waiting to be admitted. In our study, treatment with PCA was associated with a nonsignificant reduction in LOS.
Since 2013, the Hospital Readmissions Reduction Program has penalized hospitals with a higher than expected 30-day readmission rate for certain diagnoses,25 and readmission reduction has become a primary focus of quality improvement efforts and health services research.26 In previous research, patients with SCD admitted with VOC were found to be at high risk for 30-day readmission, with reported rates as high as 31%-50%.2,9 In 1 study, the primary causes of readmission were premature discharge (ie pain not well controlled), and withdrawal syndromes.9 Thus, it is important that any change in the treatment of acute VOC is examined carefully to ensure that the risk of readmission is not increased. Although in our study, the use of PCA was associated with a nonsignificant reduction in LOS and 30-day readmissions, the findings suggest that PCA may control pain more quickly and completely than standard therapy. Further study is needed to test this hypothesis.
New modalities of pain management delivery may be met by some patients with caution and distrust, especially in those who have needed high doses of bolus opioid analgesics for pain management. We are encouraged to report that the use of PCA was not associated with increased risk for discharge AMA. However, further research is needed to understand how the use of PCA affects acute pain control, patients’ experience, and patients’ satisfaction with care.
Our study has some limitations. First, this is a retrospective cohort study, so there is potential for selection bias. Encouragingly, the study groups were similar with respect to all characteristics we measured. Next, the study was performed at a single academic institution in an urban setting, so the findings may not be generalizable in all institutions. In addition, we did not directly measure patient satisfaction with PCA. Finally, we only looked at use versus nonuse of PCA. We did not address specific doses as it related to achieving pain control and therefore optimal dosing strategies remain unknown.
In conclusion, we found that the use of PCA as the treatment for a patient with SCD admitted with acute VOC was associated with a nonsignificant reduction in LOS and 30-day readmission, without an increase in discharges AMA. PCA should be considered as a valuable modality for opioid-based pain control for patients with SCD who are admitted with painful VOC.
1. Yawn BP, Buchanan GR, Afenyi-Annan AN, et al. Management of sickle cell disease
: summary of the 2014 evidence-based report by expert panel members. JAMA. 2014;312:1033–1048.
2. Brousseau DC, Owens PL, Mosso AL, et al. Acute care utilization and rehospitalizations for sickle cell disease
. JAMA. 2010;303:1288–1294.
3. Kauf TL, Coates TD, Huazhi L, et al. The cost of health care for children and adults with sickle cell disease
. Am J Hematol. 2009;84:323–327.
4. Lanzkron S, Haywood C, Segal JB, et al. Hospitalization rates and costs of care of patients with sickle-cell anemia in the state of Maryland in the era of hydroxyurea. Am J Hematol. 2006;81:927–932.
6. Steiner CA, Miller JL. Sickle Cell Disease
Patients in US Hospitals, 2004. Rockville, MD: Agency for Healthcare Research and Quality; 2006.
7. Woods K, Karrison T, Koshy M, et al. Hospital utilization patterns and costs for adult sickle cell patients in Illinois. Public Health Rep. 1997;112:44–51.
8. McNicol ED, Ferguson MC, Hudcova J. Patient controlled opioid analgesia versus non-patient controlled opioid analgesia for postoperative pain. Cochrane Database Syst Rev. 2015;6:CD003348.
9. Ballas SK, Lusardi M. Hospital readmission for adult acute sickle cell painful episodes: frequency, etiology, and prognostic significance. Am J Hematol. 2005;79:17–25.
10. Dunlop RJ, Bennett KC. Pain management for sickle cell disease
. Cochrane Database Syst Rev. 2006;2:CD003350.
11. Benjamin LJ, Dampier CD, Jacox A, et al. Guideline for the Management of Acute and Chronic Pain in Sickle Cell Disease
. Glenview, IL: American Pain Society; 1999.
12. National Institutes of Health. The Management of Sickle Cell Disease
, 4th ed. Bethesda, MD: MNH, Lung, and Blood Institute; 2002.
13. Benjamin LJ, Swinson GI, Nagel RL. Sickle cell anemia day hospital: an approach for the management of uncomplicated painful crises. Blood. 2000;95:1130–1136.
14. Gonzalez ER, Bahal N, Hansen LA, et al. Intermittent injection vs patient-controlled analgesia
for sickle cell crisis pain. Comparison in patients in the emergency department. Arch Intern Med. 1991;151:1373–1378.
15. van Beers EJ, van Tuijn CF, Nieuwkerk PT, et al. Patient-controlled analgesia
versus continuous infusion of morphine during vaso-occlusive crisis in sickle cell disease
, a randomized controlled trial. Am J Hematol. 2007;82:955–960.
16. Allen Liles E, Kirsch J, Gilchrist M, et al. Hospitalist management of vaso-occlusive pain crisis in patients with sickle cell disease
using a pathway of care. Hosp Pract. 2014;42:70–76.
17. Ibrahim SA, Kwoh CK, Krishnan E. Factors associated with patients who leave acute-care hospitals against medical advice. Am J Public Health. 2007;97:2204–2208.
18. Glasgow JM, Vaughn-Sarrazin M, Kaboli PJ. Leaving against medical advice (AMA): risk of 30-day mortality and hospital readmission. J Gen Intern Med. 2010;25:926–929.
19. Southern WN, Nahvi S, Arnsten JH. Increased risk of mortality and readmission among patients discharged against medical advice. Am J Med. 2012;125:594–602.
20. Green P, Watts D, Poole S, et al. Why patients sign out against medical advice (AMA): factors motivating patients to sign out AMA. Am J Drug Alcohol Abuse. 2004;30:489–493.
22. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613–619.
23. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383.
24. Hauck K, Zhao X. How dangerous is a day in hospital? A model of adverse events and length of stay for medical inpatients. Med Care. 2011;49:1068–1075.
26. Joynt KE, Jha AK. Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. JAMA. 2013;309:342–343.