Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Cost effectiveness analysis of intravenous ketorolac and morphine for treating pain after limb injury: double blind randomised controlled trial.

Rainer TH, Jacobs P, Ng YC et al.

Br Med J 2000;321:1247-1252. [abstract] [full-text]

Reviewed by Al Torres, MD, MS, University of Illinois College of Medicine at Peoria

Review posted November 1, 2002

I. What is being studied?:

The study objective:

To investigate the cost effectiveness of intravenous ketorolac compared with intravenous morphine in relieving pain after blunt limb injury in an accident and emergency department of a university hospital in Hong Kong.

The study design:

A double blind, randomized, controlled study and cost consequences analysis, i.e., a comparison of costs with several different outcomes. Seventy-five subjects received a 10 mg loading dose of IV ketorolac followed by 5 mg every 5 minutes up to 20 minutes (maximum 30 mg) as required and 73 subjects received a 5 mg loading dose of morphine followed by 2.5 mg every 5 minutes up to 20 minutes (maximum 15 mg) as required.

The outcomes assessed:

The primary outcome measure was a cost consequences analysis comparing the use of ketorolac with morphine; secondary outcome measures were pain relief at rest and with limb movement, adverse events, patients' satisfaction, and time spent in the emergency department. A visual analogue pain score, numbered one to ten, was used for baseline and at subsequent time intervals after the first injection up to a maximum of six hours after the analgesia administration.

II. Are the results valid?

1.Did the analysis provide a full economic comparison of the health care strategies?

Yes. A full economic comparison requires that both costs and outcomes be analyzed for each of the strategies being compared. Additional questions provided in the Users' Guide Article about assessing economic analysis (1) to assist in addressing this first question include:

A. Was a broad enough viewpoint adopted?

Yes. Cost effectiveness analyses may evaluate cost and outcomes from multiple viewpoints: the patient, the hospital, the third-party payor, or society. Since the investigators selected short-term management of limb injury pain in an emergency department, costs and outcomes were evaluated from the standpoint of the hospital. It would have been helpful for those of us in the US to know the rate of conversion to US dollars at the time of the study.

B. Were all the relevant clinical strategies compared?

No. The investigators were correct in not including a placebo control group since administration of a placebo to someone with this severity of pain would be unethical. However, the dosing regimen administered to the subjects was unique to their study. A comparison of dosing regimens used more commonly in clinical practice would have made the study results more generalizable. Including treatment of limb injury pain with oral administration of opioid analgesics would also be a relevant strategy. Alternatively, it would have been interesting to add a third comparison limb combining the two drugs. Using ketorolac and morphine in combination would likely reduce the amount of morphine delivered to reduce post-operative pain.

2. Were the costs and outcomes properly measured and valued?

A. Was clinical effectiveness established?

Yes. Effectiveness exists when a study demonstrates a treatment as it is typically administered in clinical practice performs better than another treatment, which is also administered in typical clinical fashion. The dosing regimen consisting of administering a loading dose of morphine with administration of subsequent timed doses is common practice in many emergency departments. However, it is probably less common that a non-steroidal anti-inflammatory drug is dosed in this same fashion. The study did demonstrate the efficacy of one analgesic over the other. The unique dosing regimen with a prospective economic evaluation confers high internal validity but poor external validity if this dosing is not similar to that used at other emergency departments.

B. Were costs measured accurately?

Maybe. The investigators did report the physical quantities of resources consumed by the treatments (e.g., time spent preparing the analgesic) as well as the costs attributed directly to the purchase of the medications. However, they don't explain how they obtained "the estimates for the time required to use each resource in each activity". Secretly recording the time spent performing each of the activities would probably be most accurate. The investigators reported actual costs rather than charges. Attempts at converting charges to costs using a simple conversion formula should be interpreted with caution since charges are often complex in how they are derived. Many investigators report hospital charges for the reasons aforementioned.

C. Were data on costs and outcomes appropriately integrated?

Yes. Costs and outcomes are sometimes combined into a single indicator (e.g., quality-adjusted life year, QALY) for comparison. Combining cost and the outcome variables was not necessary since the incremental costs and other outcome variables (e.g., mean difference in pain score change) were easy to compare individually between both groups.

3. Was appropriate allowance made for uncertainties in the analysis?

No. Sensitivity analyses are often performed in cost effectiveness analyses to adjust the results for variables that are estimated. Although the investigators described conducting a sensitivity analysis of cost measures with regards to the observation period, they didn't allow for uncertainties in personnel time, i.e. variance in drug prep time, direct nursing care, physician time, etc. They didn't allow for uncertainties in inpatient costs. They also didn't allow any adjustments in drug costs, which could change with the next pharmacy contract. A thorough and detailed cost analysis usually provides multiple cost outcomes after allowing for variation in several of the variables. The results would have been more convincing if the investigators had assumed that the patients discharged prior to the end of the six-hour observation period were failures and they still demonstrated a significant difference in the treatment group.

4. Are estimates of costs and outcomes related to the baseline risk in the treatment population?

Yes. All patients enrolled were relatively healthy with the exception of the traumatic limb injury. Patients with chronic diseases/conditions were excluded. Patients with chronic pain or a tolerance to analgesics would have influenced the success of the dosing regimens. Also, the impact of complication rates on costs and the other outcomes would have been higher in patients with chronic diseases/conditions.

III. What are the results?

1.What were the incremental costs and outcomes of each strategy?

The investigators did report the costs related to the analgesics and the human resources consumed (i.e., upfront costs) as well as the reattendance or admission costs (i.e., downstream costs). The average cost per person, excluding admissions not related adverse events, was $HK 43.60 for those in the ketorolac group and $HK 228.80 for those in the morphine group (p < 0.0001). The majority of the differences in costs are attributable to management of adverse events (i.e., administration of metoclopramide and naloxone, doctors' and nurses' time to manage adverse effects, and daily admission costs for adverse effects). Twenty-two patients in the morphine group received metoclopramide and/or naloxone where as none of the patients in the ketorolac group did. Three patients were admitted for adverse events in the morphine group where as none of the patients in the ketorolac group were.

2. Do incremental costs and outcomes differ between subgroups?

No. The following table lists the variables significantly different between the ketorolac and morphine groups. The investigators did not present any data on cost savings for each admission adverted, or costs per each unit improvement in pain scores.

Variable	Ketorolac group	Morphine group	Median diff (95% CI)	P value
Median min (interquartile range) to prepare analgesia	5.0 (5.0 - 10.0)	10.0 (5.5 - 12.5)	2.0 (0 to 5.0)	0.0002
Interval between receiving analgesic and leaving ED	115.0 (75.0 to 149.0)	130.0 (95.0 - 170.0)	20.0 (4.0 - 39.0)	0.02
Total no. (%) of subjects with adverse events	4 (5)	65 (89)	NA	< 0.0001
Median drop (95% CI) in pain score/hr with activity	1.09 (1.05, 2.02)	0.87 (0.84 to 1.06)	NA	0.003
Median (interquartile range) satisfaction with analgesia	6.0 (5.0 - 6.0)	5.0 (4.0 - 6.0)	1.0 (1.0 - 2.0)	< 0.0001

The odds of experiencing adverse events were 144.2 (95 CI 41.5, 501.6) times more likely with morphine than with ketorolac. The mean cost per person was $HK44 in the ketorolac group and $HK229 in the morphine group (P < 0.0001).

3. How much does allowance for uncertainty change the results?

Examining the lower limit of the 95% CI for some the significant differences reveal how clinically small the differences may actually be. For example, the lower limit of the 95% CI for the median difference in minutes to prepare the analgesics was zero. Despite the significant P value for this variable, the 95% CI reveals the possibility that there may not be any difference in time to prepare the two analgesics.

IV. Will the results help me in caring for my patients?

1. Are the treatment benefits worth the harms and costs?

The investigators demonstrated that ketorolac was as effective as morphine and less costly because of its lower incidence of adverse effects and less time to prepare and administer the former. The investigators point out the increased time needed to prepare and administer morphine was due to the standard procedure for the nurses to double check the opioid dose with another nurse. According to the 3 x 3 matrix for decision making in the Users' Guide to Cost Effective Analysis (2), ketorolac has a weak dominance over morphine in the treatment of pain associated with limb injury. The matrix plots effectiveness vs. costs so that a more effective treatment at less cost is strongly dominant to accept the treatment and a less effective treatment at higher cost is strongly dominant to reject the treatment.

2. Could my patients expect similar health outcomes?

No. A dosing regimen of ketorolac consisting of a loading dose followed by subsequent doses at minimal intervals have not been adequately tested for safety and efficacy in children. Ketorolac is usually administered at the maximum dose every six to eight hours, but morphine is often titrated to effect. It is reasonable to expect that if ketorolac and morphine were administered in equipotent analgesic doses that morphine would result in more adverse effects. There is adequate evidence demonstrating the need for less opioid if children with postoperative pain receive ketorolac but no one has done a direct comparison of these two analgesics without crossover. Again, a third experimental arm combining the two drugs to determine if indeed lower doses of morphine could have been used with fewer side effects would have been interesting.

3. Could I expect similar costs?

Although ketorolac does not cost 2.5 times as much as morphine at our institution ($15 USD for ketorolac vs. $11 USD for morphine), the time spent preparing and administering morphine would likely be significantly longer than the time needed to do the same with ketorolac. Also, as stated above, it is reasonable to assume the children in the morphine group would have more adverse events also increasing the costs. But alas, further study in children is needed.

References

1. Drummond MF, Richardson WF, O'Brien BJ, et al. User's guide to the medical literature. XIII. How to use an article on economic analysis of clinical practice. A. Are the results of the study valid. JAMA 1997; 277:1552-7 [full-text]
2. O'Brien BJ, Heyland D, Richardson SW, et al. User's guide to the medical literature. XIII. How to use an article on economic analysis of clinical practice. B. What are the results and will they help me in caring for my patients? JAMA 1997; 277:1802-6 [full-text]

 

Comments

Back to the EB Journal Club Index