David Daniel knows his son needs help.

The 8-year-old spent first grade in remote learning and several weeks of second grade in quarantine. The best way to catch him up, research suggests, is to tutor him several times a week during school.

But his Indianapolis school offers Saturday or after-school tutoring—programs that don’t work for Daniel, a single father. The upshot is his son, now in third grade, isn’t getting the tutoring he needs.

“I want him to have the help,” Daniel said. Without it, “next year is going to be really hard on him.”

As America’s schools confront dramatic learning setbacks caused by the pandemic, experts have held up intensive tutoring as the single best antidote. Yet even as schools wield billions of dollars in federal COVID relief, a small fraction of students have received school tutoring, according to a survey of the nation’s largest districts by the nonprofit news organization Chalkbeat and The Associated Press.

In eight of 12 school systems that provided data, less than 10 percent of students received any type of district tutoring this fall. To compare, in a federal survey, school officials said half of all U.S. students started this school year behind grade level in at least one subject.

A new tutoring corps in Chicago has served about 3 percent of students, officials said. The figure was less than 1 percent in three districts: Georgia’s Gwinnett County, Florida’s Miami-Dade County, and Philadelphia, where the district reported only about 800 students were tutored. In those three systems alone, there were more than 600,000 students who spent no time in a district tutoring program this fall.

The startlingly low tutoring figures point to several problems. Some parents said they didn’t know tutoring was available or didn’t think their children needed it. Some school systems have struggled to hire tutors. Other school systems said the small tutoring programs were intentional, part of an effort to focus on students with the greatest needs.

Whatever the reason, the impact is clear: At a crucial time for students’ recovery, millions of children have not received the academic equivalent of powerful medication.

“It works, it’s effective, it gets students to improve in their learning and catch up,” said Amie Rapaport, a University of Southern California researcher who has analyzed students’ access to intensive tutoring. “So why isn’t it reaching them?”

The Indianapolis school district last year launched two tutoring programs that connect students with certified teachers over video. One is available to all students after school, while the other is offered during the day at certain low-performing schools.

District officials say a trial run boosted student test scores. Parents give it high marks.

“The progress that he made in just a couple months last semester working with his tutor was kind of far beyond what he was grasping and doing at school,” said Jessica Blalack, whose 7-year-old, Phoenix, opted in to after-school tutoring.

Still, the two programs combined served only about 3,200 students last fall, or roughly 17 percent of students in district-run schools. Two additional tutoring programs operate at a handful of schools.

Only 35 percent of the students who registered for after-school tutoring last fall attended more than one session, according to district data.

Indianapolis Public Schools spokesperson Marc Ransford said the district is working to improve attendance and hopes to enroll more students in tutoring next school year. It’s also trying to accelerate student learning in other ways, including with a new curriculum and summer school.

Nationwide, schools report that about 10 percent of students are receiving “high-dosage” tutoring multiple days a week, according to a federal survey from December. The real number could be even lower: Just 2 percent of U.S. households say their children are getting that kind of intensive tutoring, according to the USC analysis of a different nationally representative survey.

Schools trying to ramp up tutoring have run into roadblocks, including staffing and scheduling. Experts say tutoring is most effective when provided three times a week for at least 30 minutes during school hours. Offering after-school or weekend tutoring is simpler, but turnout is often low.

Harrison Tran, a 10th grader in Savannah, Georgia, struggled to make sense of algebra during remote learning. Last year, his high school offered after-school help. But that wasn’t feasible for Harrison, who lives 30 minutes from school and couldn’t afford to miss his ride home.

Without tutoring help, he started this school year with gaps in his learning.

“When I got into my Algebra II class, I was entirely lost,” he said.

Relatively low family interest has been another challenge. Though test scores plunged during the pandemic, many parents do not believe their children experienced learning loss, or simply are unaware. The disconnect makes it more important to offer tutoring during school, experts say.

“Parents just aren’t as concerned as we need them to be,” said USC education professor Morgan Polikoff, “if we’re going to have to rely on parents opting their kids into interventions.”

Even when students want help, some have been let down.

In Maryland’s Montgomery County, 12th grader Talia Bradley recently sought calculus help from a virtual tutoring company hired by the district. But the problem she was struggling with also stumped the tutor. After an hour of trying to sort it out, Talia walked away frustrated.

“My daughter was no farther along,” said Leah Bradley, her mother. “Having an option for online tutoring makes sense, but it can’t be the primary option if you’re looking for good results.”

Repeated in-person tutoring tends to be more effective than on-demand online help, but it’s also harder to manage. District rules add complexity, with safeguards like tutor background checks and vendor bidding rules slowing the process.

In Wake County, North Carolina, the school district began planning a reading tutoring program last summer. The program did not launch until November, and district officials last month said volunteers are tutoring fewer than 140 students—far fewer than the 1,000 students the program was designed to reach.

“We’re always looking to serve more students,” said Amy Mattingly, director of K-12 programs at Helps Education Fund, the nonprofit managing that program and another serving about 400 students. But, she added, it’s important to “see what’s working and make tweaks before trying to scale up.”

Some districts defended their participation numbers, saying tutoring is most effective when targeted.

In Georgia’s Fulton County, 3 percent of the district’s 90,000 students participated in tutoring programs this fall. Most of the tutoring was offered by paraprofessionals during the school day, with one hired to give intense support in each elementary school.

The district says time and staffing limit how many students can get frequent, intensive tutoring.

“We don’t want to water it down, because then you don’t get the impact that the research says is beneficial for kids,” said Cliff Jones, chief academic officer for the system.

Others worry too few are getting the help they need even as programs continue to grow.

This school year, about 3,500 students are getting reading tutoring from the North Carolina Education Corps. Meanwhile, in fourth grade alone, more than 41,000 students statewide scored in the bottom level on a national reading test last year.

“Who we are serving,” said Laura Bilbro-Berry, the program’s senior director, “is just a drop in the bucket.”

This article contains information from The Associated Press.