It's taking too long to get vaccine doses from refrigerators into arms
The emergence of dangerous new coronavirus variants underscores the urgency of administering available doses of vaccine quickly. To date, however, that effort has fallen short. Too many jurisdictions appear to be allowing too much vaccine to sit idle for too long before getting it into the arms of residents. The data in this posting show which jurisdictions have been putting their doses to work most quickly and which have been lagging behind.
Speed of vaccine administration does not tell the whole story. A jurisdiction may perform well on this metric but abysmally on other dimensions. For example, a jurisdiction prioritizing speed of administration may meet its goal by going first to easy-to-reach populations, with adverse consequences for equity if minority or immigrant communities are harder to reach. Another jurisdiction may do well on administering vaccine but poorly overall, because it encourages disregard for mask wearing, social distancing, and other good public health practices. Although speed is not the only consideration, it is an important one, so it is worth measuring well.
Thus far, a common way to assess the speed of vaccine administration by state and local authorities has been to calculate the number of doses administered as a percent of doses delivered to these governments by the federal government. In the early going, this has been a useful approach. But as volumes ramp up, it will become less meaningful. To see why, consider that by January 14, the federal government had delivered 30 million doses of vaccine to various jurisdictions around the country. Of that amount, only 11 million had been administered. The other 19 million doses sat in refrigerators, not yet put to work. Many analysts decried the fact that only 36 percent of available vaccine had been administered. But suppose we look down the road a few months, to when five times as many doses—150 million—have been delivered. Even if the same large number of doses (19 million) were to be sitting idle, the number of doses administered as a percent of doses delivered would be a more impressive 87 percent. State and local performance would remain unsatisfactory, but the usual metric would seem to point to a more positive conclusion.
A better performance measure would be the number of days from when the state and local authorities take possession of the vaccine to when they get it into people’s arms. The accompanying figure and associated table implement this approach. On this metric, the best performing state as of late January was North Dakota. As of January 30, they had reported having administered 91,585 doses. As of January 26, they had received 86,750 doses; as of January 27, they had received 96,425 doses. Therefore, according to this data source, as of the 30th, they had administered as many doses as had been delivered four or more days earlier and must have been administering some doses delivered three or fewer days earlier. Accordingly, North Dakota is scored as having administered, as of January 30, all doses received four or more days earlier. West Virginia had done almost as well: As of the 30th, they had administered as many doses as had been delivered five or more days earlier and must have been administering some doses delivered only four or fewer days earlier.
At the other end of the spectrum, 22 jurisdictions—including 13 states—had not, as of January 30, administered as many doses as they had already received by January 12, 18 days earlier. Many different factors may help explain the delay, but 18 days is too long to wait.
The path ahead will not be easy, but it is essential that available doses be administered quickly, equitably, with careful attention to saving lives, and in a manner that honors our front-line medical workers. We have a long way to go.
|How many days from delivery did it take jurisdictions to administer COVID-19 vaccines?|
|Number of days prior to January 30 that as many vaccines as received were administered, by jurisdiction|
|0 days||Bureau of Prisons|
|4 days||North Dakota|
|5 days||West Virginia|
|8 days||New Mexico, South Dakota|
|9 days||District of Columbia|
|11 days||Colorado, Connecticut, Delaware, Kentucky, Louisiana, Michigan, Montana, Nevada, New Jersey, Oregon, South Carolina, Texas, Utah, Wisconsin|
|16 days||Mississippi, Oklahoma, Rhode Island|
|17 days||Department of Defense, Florida, Illinois, Indiana, North Carolina, Pennsylvania, Washington|
|18 days||Arkansas, Iowa, Kansas, Massachusetts, Minnesota, New Hampshire, New York, Ohio, Tennessee, Veterans Health, Wyoming|
|More than 18 days||Alabama, Alaska, American Samoa, Arizona, California, Federated States of Micronesia, Georgia, Guam, Hawaii, Idaho, Indian Health Svc, Maine, Marshall Islands, Maryland, Missouri, Nebraska, Northern Mariana Islands, Puerto Rico, Republic of Palau, Vermont, Virgin Islands, Virginia|
|Sources: Author's calculations based on data from the Centers for Disease Control and Prevention, by way of “Our World In Data,” downloaded January 31, 2021.|
1. As shown in the table, the best performing jurisdiction overall was not a state but rather the Federal Bureau of Prisons. In fact, the Bureau of Prisons reported having administered more doses as of January 30 than the federal government had delivered to it. This apparent anomaly may reflect the fact that in some cases, it has been possible to extract more doses per vial than the manufacturer claimed credit for.
2. The calculations reported in this post do not take account of doses lost to spoilage or other similar causes.
3. The data shown here should not be affected by the federal government’s initial intention to withhold enough supply to provide second doses. Assuming those doses were warehoused at the federal level, they would not show up in deliveries to the state and other jurisdictions, and therefore should not have been an impediment to speedy administration.