Vaccine Presentation in the USA: Economics of Prefilled Syringes versus Multidose Vials for Influenza Vaccination

Claudia C Pereira; David Bishai


Expert Rev Vaccines. 2010;9(11):1343-1349. 

In This Article

Efficiency & Costs: Baltimore Study


The efficiency, costs and safety practices related to influenza vaccination were evaluated in a time-motion study. Seven practices in Baltimore (MD, USA) and the metropolitan area participated in the study. The practices included hospitals, clinics and private medical centers. The sites varied in their location; three were urban and four suburban. Five practices provided both seasonal and H1N1 influenza vaccines, whereas two only provided the seasonal influenza vaccine. We observed 31 immunization professionals (Table 1), all of whom were registered nurses with various years of experience, ranging from 3 to 39 years (mean: 18.1; SD: 11.7). The practices sampled were selected based on their willingness to participate in the study and permission was obtained from each healthcare provider to have their immunization-related activities observed and timed. Observations and data collection took place between October 2009 and March 2010.

Data Collection

We collected time and cost data on all activities related to immunization with MDVs and PFSs for both seasonal and H1N1 influenza vaccines. The time spent by immunization professionals on their vaccination-related activities was recorded using a stopwatch. We observed tasks in the vaccine preparation process, from the time of removal of the vial or the syringe package from the refrigerator until the time it was ready to be administered to a patient. Often during the observations it was not possible to observe all tasks involved in one vaccine preparation process as a whole, as there was no standardized way of executing the tasks. For instance, sometimes nurses attached needles to syringes in bulk, so they were readily available when needed. That may have happened hours or days before they actually drew vaccine from a vial, so that the observer was not able to observe that step, but could observe the other steps of the sequence. We assumed that the time to actually inject the patient would be the same for both methods, so we did not measure the time it took to inject. Our time measurements end from the moment the vaccine was considered ready to inject.

Cost Measurement

Healthcare professionals' salaries were obtained through the US Bureau of Labor Statistics website, statistics for the year 2009.[11] We utilized mean wages for registered nurses in the state of Maryland, USA.

Monthly storage costs were estimated by obtaining the average market price of medical-use refrigerators of 4 cubic feet capacity, assuming a 10-year useful life and a discount rate of 5%. Consumer price inflators were used to adjust prices for inflation. The average volume for a package of five doses of PFSs was 52.6 cm3 and of ten doses of MDVs or one vial was 5.4 cm3.

Disposal/waste cost difference was assumed to be negligible, as we pressumed that hospitals and clinics already have a contracted waste service and the choice of MDVs or PFSs would not lead to additional waste disposal collection visits.

The prices for syringes, needles, alcohol pads and gloves for the brands and models we observed were obtained from the 2009 Red Book™ Drug Reference.[12] The prices for vaccines were obtained from the IMS MIDAS database.[108]

In order to account for uncertainty and to take into account the relatively small number of practices we investigated, we performed a set of one-way sensitivity analyses, using the lowest and highest costs we identified in the 2009 Red Book for vaccines, needles, gloves and alcohol. For storage we calculated the lower bound by multiplying the smallest storage unit cost by 0.8, and likewise we multiplied the largest storage cost for PFSs by 1.2. In a similar fashion, for the lower bound of healthcare worker wages, we multiplied the national average salary of a medical assistant by 0.8. For the upper bound, we used the mean salary for a registered nurse.

Statistical Analyses: Time Use Estimation

All analyses were made using STATA SE 10.[13] The process to prepare a vaccine for injection is composed of several individual subtasks. Some subtasks were timed individually. Most were timed as part of a continuum of sequential subtasks. Stop watch episodes commonly represented more than one single subtask, but the duration of each episode was always accompanied by data on which subtasks had occurred in the interval. In our dataset we had up to seven subtasks represented in one single time measurement. The dataset was arranged so that the length of each measured episode was attached to a list of up to seven subtasks that occurred during that interval. Getting a vaccine ready from MDVs could take up to 16 subtasks and up to ten for PFSs (Figure 1).

Figure 1.

Estimated time in seconds for tasks associated with multidose vial and prefilled syringe use.
MDV: Multidose vial; PFS: Prefilled syringe.

To estimate time per subtask we used random-effects regression models to take into account the effects of healthcare professionals. The model assumed clinic-specific and worker-specific random effects. We were able to estimate individual weighted average times for each of the 21 subtasks that could possibly be performed. From our observations and from the literature, we listed all practice steps to vaccinate using MDVs and PFSs, although sometimes none of these steps were performed by a nurse. Finally, we estimated total time required to complete all subtasks for either MDVs or PFSs. The coefficients in the models were interpreted as the incremental number of seconds required to conduct each corresponding subtask. After estimating the average time required for each of the subtasks we observed, we estimated the total time to conduct the entire sequence of subtasks (Figure 1). We conducted that separately for vial-only tasks (13 tasks), common tasks for both MDVs and PFSs (six tasks) and PFS-only tasks (five tasks).


We observed a total of 555 multitasking activities performed by 31 different registered nurses. We estimated that the total time in seconds to get a vaccine ready to be administered per patient was 87 s for MDVs and 49.71 s for PFSs, a difference of 37.29 s (Figure 1).

After adding all relevant administration costs, which included nurses' time, syringes for MDVs only, needles, storage, alcohol and gloves per 1000 injections, the total administration costs in 2009 were US$8596 for MDVs and US$8920.21 for PFSs, a difference of US$324.21 (Table 2) or US$0.32 per dose administered. These numbers exclude the acquisition cost of the vaccine. Our estimates thus suggest a 'breakeven' PFS price differential of US$0.32. At this price, the savings in vaccination administration costs would exactly offset the higher PFS purchase price.

We conducted a series of univariate sensitivity analyses to explore the impact of varying all costs that were common to MDVs and PFSs (Figure 2). Our results indicate that vaccine acquisition price is the main source of uncertainty in the analyses. The total administration costs varied from US$7456 to US$10,010 in 2009 dollars.

Figure 2.

Tornado diagram of univariate analyses. This tornado diagram shows the degree to which uncertainty in prices affects estimates.

A number of clinical practice adaptations during vaccine preparation were observed and documented. At times, for instance, the same alcohol swab was repeatedly used to sterilize the vial, without being exchanged or receiving more alcohol. This was especially true when nurses were predrawing a batch of vaccines at once in preparation for a big influenza clinic day, so they had to act quickly. The observer noted instances where nurses pooled vaccine remainders from multiple spent vials in order to assemble a full dose, which does not adhere to safe injection practices.

In many clinics, vaccines were predrawn many hours in advance to expedite the vaccination process, especially in influenza clinics that need to minimize patient waiting time. With predrawing there was potential confusion about vaccine lot number, since there was no place on the syringe to write the lot number. In instances of predrawing, we frequently observed incorrect forecasts of the number of vaccines that would be needed. Leftover vaccines that were predrawn from vials would remain in the plastic syringe for the following day.

When vaccines were predrawn, nurses often relied on memory for entering vaccine lot numbers in patient records. A very common practice was to memorize one relevant lot number for a day or session and enter the memorized number on the patient's paperwork, without referencing the original packaging. Some nurses wrote the vial lot numbers on pieces of paper that they could later refer to and some even wrote the number on their skin, as a way of being able to refer to it at any time. During one session the lot number had changed between the syringes for that session and a nurse was still writing the previous lot number.


Although MDVs require less cold storage and may have lower acquisition costs, their use imposes higher staff time burdens and higher task complexity. Practices related to predrawing vaccine into unmarked plastic syringes were observed that could impact patient safety and reduce vaccine potency. PFS packaging makes them bulkier to store but none of the clinics observed had cold storage constraints that would impede their ability to use PFSs. By the same token, none of the clinics were at a point where they could reduce total clinic staff if they could save 37.29 s per patient vaccinated based on the patient volumes the days we observed.

Nevertheless, in the broad scheme of things, the 37.29-s gap between MDVs and PFSs may have important implications. Considering the American birth cohort of 4 million infants, with each infant typically encountering approximately 15 shots before 1 year of age, the use of PFSs would save 621,500 person-hours worth US$22.2 million in healthcare worker time in 1 year of infant vaccination. In a pandemic situation where 300 million Americans would require vaccination, PFSs would save 3.12 million h, worth US$111.1 million in healthcare worker time. The economic decision rests on whether there are better uses for the precious resource of healthcare worker time, especially of nurses.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: