NJ School Nurses: A Data-Driven View

UPDATE: I forgot to include this: I left "special services" districts out of the analysis. These districts, which have specialized needs due to their student populations, shouldn't be compared to others, and I have reason to suspect the data isn't set up to make comparisons anyway.


In my last post, I note that one in five U.S. public schools has no nursing coverage.

Given the COVID-19 outbreak, this is disturbing. School nurses are on the front lines of pediatric care, and a critical part of any response this nation will have to a pandemic. Every American child deserves access to a qualified, well-trained school nurse.

Ed Fuller at Penn State (as big an edu-data geek as yours truly) has tweeted out some graphs showing the distribution of nurses in Pennsylvania schools.

Schools serving our children most in need of counselors & nurses are least likely to have access to them in PA. We must change that! pic.twitter.com/SVCXlH3UUo

— Ed Fuller (@EdFuller_PSU) March 8, 2020

What do we see? In PA, nursing coverage decreases when poverty rises, in elementary, middle, and high schools. In other words, at a time when the coronavirus threat is looming, Pennsylvania's neediest children are less likely to get the medical care they need at school.

Given all this, I thought it would be useful to take a look at how New Jersey's school nursing situation stands. I've got the particulars about the data below, but for now: I used two different data sources from the NJ Department of Education, and got very similar results from each.

My unit of analysis is a school district, rather than a school. New Jersey has a lot of small districts with only two or three buildings, often feeding into a regional high school with several other small districts. It's possible these buildings share a nurse, but the time between buildings isn't correctly divided in the records. With my method, a nurse assigned to any building in the district counts for the entire district.

My measure for these graphs is "students per nurse": in other words, how many students, on average, are there in a district for every nurse in the records? Let's start by looking at how districts with different levels of student economic disadvantage compare.

I've divided all NJ districts into five groups, ranging from those with the fewest students who qualify for free lunch (a proxy measure of economic disadvantage) to those with the most. I was, to be honest, surprised to see this: in New Jersey, school nursing care is distributed quite equally. The student load for nurses in the most disadvantaged districts is, on average, close to the load for nurses in the least disadvantaged districts.

Of course, an equal distribution isn't necessarily "fair." We can safely assume many of the students in high-FRPL districts don't have adequate medical outside of school. If we want children to have an equal chance for school success, we'd want to see a lower student load in the most disadvantaged districts.

That said: at least the coverage isn't flipped, like in PA. And the average load is under the recommended ratio of 750:1, as suggested by the American Academy of Pediatrics. So, a good start from New Jersey... but we probably can and should do better.

Let's look at distribution by race next:

Here I'm combining Black and Hispanic students, but I did run the numbers separately for each group, and the results are similar. The distribution of nurses is quite equal across race and ethnicity in New Jersey; there's really no evidence of meaningful bias. Given the correlation of race and income, however, we would like to see smaller student loads for nurses in districts with higher concentrations of students of color if we want to achieve equal educational opportunity. But, again, we're not Pennsylvania, and the student loads, on average, are below recommended guidelines.

New Jersey has many different types of districts: smaller, larger, some only K through 8, some K through 12, some regional high schools, etc. Vocational-technical districts and charter schools are also considered their own districts. So how does nurse coverage vary across these different district types?

Now we're getting some variation. As I said, NJ has a many small K-8 only districts: these districts, it appears, hire nurses for their schools, even if the student populations are small. K-8 districts with under 400 students have a student:nurse ration of 238-1; but K-8 districts with more than 750 students have a ratio of 435-1.

A perpetual topic of conversation in New Jersey education policy is how we have too many undersized districts that are inefficiently small. I think the fiscal pressure this creates has often been exaggerated; nonetheless, there are some real inefficiencies that come from having undersized school districts, and this may be one of them.

The regional high schools (7-12/9-12) have, on average, much higher student loads for their nurses. Is this a problem? I don't know... intuitively, it makes sense to me that we'd want more nursing services for younger children, who are less able to participate in monitoring their own health care. But we should note the vo-techs, which are basically high schools, have a considerably lower nurse-student ratio than the regional high schools. If it's good for those kids, why not others?

Which brings us to the charter schools, with much higher student loads for their nurses. The majority of the charter population in New Jersey consists of K-8 students, but there are a significant number of high school students in the charters. Still, even if we accept the premise that older students don't need as much nursing coverage as younger ones, the high student load for charter nurses stands out.

What's driving the difference? According to state records, there are many charter school students who have no nursing care available.

29 percent of New Jersey charter school students do not have a school nurse in state records listed as on staff for their school. Now, we shouldn't immediately jump to a conclusion that they don't have access to nursing care. Maybe the charter contracts out its nurses and doesn't list them as staff. It's worth noting, however, that the "preparedness review" process for charters doesn't explicitly call for a review of nursing services.

As Bruce Baker has noted, charter schools are usually small and, therefore, more likely to induce inefficiencies. We can actually see this here: very small K-6 & K-8 districts are more likely to not have a school nurse in the records than larger districts. Is the size of these districts, and the charters, precluding students from receiving nursing care in their schools?

Also: two-thirds of NJ's charters are located in the boundaries of the least-affluent school districts. It could be that the lack of charter school nurses is inflating the average student-nurse ratios for schools with larger concentrations of disadvantaged students (more on this later).

We should know all this for sure: part of the oversight process for charter schools (and all schools) should include documenting access to school nurses. All students deserve nursing care in their schools, and school boards -- whether for districts, vo-techs, or charters -- must be held accountable for whether nurses are available for their students.


ABOUT THE DATA:

I used the NJDOE staffing files for these graphs, linked to enrollment and other files, for this analysis. I checked my results against the NJDOE's School Performance Reports, which also track student-nurse ratios at the district level. The results from the two analyses were very similar.

If there was no record of a nurse on staff, I assumed any students in that district/charter had no nursing care. The averages across race/ethnicity, FRPL, and district type reflect this assumption.