Clinical Trial Enrollment Tracking: A Forecasting Instrument

At a glance

• Enrollment tracking is not a “who signed up” list. Done right, it is a forecasting instrument: actual enrollment plotted against the projected curve, read early enough to act.
• Three metrics turn a log into an instrument: the cumulative projected-vs-actual curve, the screen-failure rate, and the screening-to-enrollment ratio.
• The point of the curve is variance: a widening gap between projected and actual is your early warning that a site or the study is lagging, while there is still time to intervene.
• A simple Excel tracker can compute all of this; the article describes the sheet so you can build it.
• Spreadsheets break for enrollment tracking when sites multiply and updates lag, which is the cue to move to a tool that updates the curve live.

Most “enrollment tracking” templates give you a place to record who enrolled and when, a log. What they never teach is how to read it: how to tell, in week six, that you are quietly falling behind, and which site is dragging. Enrollment is the single biggest driver of trial timelines and cost, so a tracker that only records the past is a missed opportunity. A tracker that forecasts is a management tool.

This guide reframes enrollment tracking as a forecasting instrument. It defines the metrics that matter, shows how to construct the tracker, explains how to read a lagging curve, and notes where Excel stops keeping up. It stays on tracking, not recruitment strategy or the broader timeline (those are separate topics).

Enrollment tracking vs. an enrollment log

First, a distinction that trips people up. An enrollment log and enrollment tracking are not the same thing, and conflating them is why so many teams think they are “tracking” when they are only recording.

• An enrollment log is a source record: the list of who was screened and enrolled, when, and at which site. It is a factual register, part of the trial’s documentation, and it answers “what happened.”
• Enrollment tracking is a management metric built on top of that log: the projection, the curve, the rates, the variance. It answers “are we on pace, and if not, where and why.”

You need both, but they serve different masters. The log is for the record; the tracking is for the decision. A page that hands you a blank log has given you the raw material for tracking, not tracking itself. The rest of this guide is about turning the log into the instrument.

Why enrollment tracking is oversight, not bookkeeping

Enrollment pace determines whether a trial finishes on time and on budget, and lagging enrollment is one of the most common reasons trials slip. Watching it is part of running the trial well: ICH E6(R3) frames trial processes as proportionate to the risks inherent in the trial and to the importance of the information collected, with quality built in by identifying prospectively the factors critical to the reliability of the results (§ 6.2, § 7.1). Enrollment is squarely one of those critical factors, because a trial that cannot recruit cannot produce reliable results on time. Tracking it as a forecast, not a tally, is how you manage that risk while you still can.

The metrics that turn a log into an instrument

The projected-vs-actual enrollment curve

The core artifact is a cumulative curve: for each time period, the number of participants you projected to have enrolled, plotted against the number you actually have. The projection comes from your enrollment plan (sites times expected rate per site per month, ramped for activation). The actual is your real cumulative count. The two lines together are the instrument; the gap between them is the signal.

Screen-failure rate

Not everyone screened enrolls. The screen-failure rate is the proportion of screened candidates who fail screening (do not meet eligibility). A high or rising screen-failure rate tells you your screening funnel is leaky, perhaps eligibility criteria are too tight, or a site is screening the wrong candidates, and it explains why enrollment lags even when screening looks busy.

Screening-to-enrollment ratio

Closely related: how many people you must screen to enroll one participant. If the ratio is 3:1, you need to screen three candidates per enrollment; if it drifts to 5:1, your sites must screen far more to hit the same target, which has real workload and timeline implications. Tracking this ratio lets you forecast the screening volume needed to hit your enrollment goal.

Per-site enrollment rate and time-to-first-patient

Two more metrics turn the aggregate into something you can act on site by site. The per-site enrollment rate (participants enrolled per site per month) is what lets you compare sites against each other and against the rate your projection assumed; a site running at half its assumed rate is the one to call. Time-to-first-patient (days from site activation to that site’s first enrollment) is an early predictor: a site slow to enroll its first participant is usually slow throughout, so a long time-to-first-patient is a warning you get weeks before the cumulative curve confirms it. Track both per site, not just at the study level, because the study average almost always hides one or two laggards.

Constructing the tracker (the Excel sheet)

A workable enrollment tracker is a few connected parts:

Sheet / section	Contents
Projection	Period, projected cumulative enrollment (from sites x rate, ramped)
Actual log	Per-participant: site, screening date, screen result, enrollment date
Site rollup	Per site: screened, screen-failed, enrolled, rate vs. target
Metrics	Cumulative actual, screen-failure rate, screening-to-enrollment ratio
Curve	Chart: projected vs. actual cumulative over time

Build it so the actual log feeds the rollup and the metrics automatically, and the chart plots projected against actual. Then the tracker computes the curve and the ratios for you, rather than you eyeballing a list.

Reading a lagging curve

The value is in the reading. A few patterns and what they tell you:

• Actual tracking below projected, gap widening. You are falling behind, and the trend is worsening. Act now, before the gap is unrecoverable.
• Overall on track but one site flat. A specific site is the problem; the aggregate hides it. The site rollup is what surfaces this, which is why per-site tracking matters.
• Enrollment lagging but screening high, with a rising screen-failure rate. The problem is the funnel, not effort: candidates are being screened but failing eligibility. Investigate criteria or screening targeting.
• Screening-to-enrollment ratio climbing. You need more screening volume per enrollment than planned; recalibrate the projection and the site workload.

The discipline is to read the curve early and often. The whole reason to forecast is to catch a lagging site in week six, when adding a backup site or fixing a screening problem can still recover the timeline, rather than discovering it at the planned enrollment deadline.

A short worked example makes the arithmetic concrete. Say you plan four sites, each assumed to enroll two participants per month after a one-month activation ramp, toward a target of 80 participants. By month four your projection is roughly 4 sites x 2 per month x 3 active months, about 24 enrolled. If your actual is 15, you are at 63 percent of projection with a nine-participant gap that is widening, the trend that demands action now. Drop to the site rollup and you might find three sites near plan and one at zero: that site’s time-to-first-patient has already blown past a month, so it, not the study as a whole, is the problem. If instead every site is screening heavily but the screen-failure rate has climbed from 40 to 65 percent, the gap is a funnel problem, not an effort problem, and the fix is eligibility or screening targeting rather than another site. The numbers point at the intervention; that is the whole value of tracking them.

Where Excel breaks

An Excel enrollment tracker works well for a small study with a few sites. It breaks when sites multiply and the data comes from many hands: updates lag, the curve is only as current as the last manual refresh, concurrent edits collide, and the per-site rollups drift. For a multi-site study, an enrollment tracker that is a week stale is not an early-warning instrument, it is a late-warning one.

The fix is a tool that keeps enrollment current and the curve live. TrialTrack is one option for small teams: it tracks enrollment by site as part of its clinical project management, so the picture stays current without a manual refresh (it is a coordination tool, not an EDC, and it does not make anyone compliant). For a lean team running several sites, live enrollment tracking is the difference between forecasting and bookkeeping.

Frequently asked questions

What is clinical trial enrollment tracking? Monitoring how enrollment is progressing against plan, ideally as a projected-vs-actual cumulative curve plus screen-failure and screening-to-enrollment metrics, so you can act on a lagging trend early.

What metrics matter? The cumulative projected-vs-actual enrollment curve, the screen-failure rate, and the screening-to-enrollment ratio. Together they tell you not just where you are but why and what to do.

How do you read a lagging enrollment curve? Watch the gap between projected and actual; a widening gap is an early warning. Use per-site rollups to find which site is lagging, and the screen-failure rate and screening ratio to tell whether the problem is effort or the eligibility funnel.

Why track screen-failure rate and screening-to-enrollment ratio? Because enrollment can lag even when screening is busy. These metrics reveal a leaky funnel and tell you how much screening volume you need to hit the target.

When does an Excel enrollment tracker stop working? When multiple sites and many editors make it lag and drift, so the curve is no longer current enough to give early warning. That is the cue to move to a tool that updates live.

The bottom line

Treat enrollment tracking as a forecasting instrument, not a sign-up log. Plot actual enrollment against the projected curve, compute screen-failure and screening-to-enrollment ratios, and read the variance early enough to fix a lagging site while you still can. Build it in Excel to start, and graduate to a live tool when multiple sites make a manual tracker too stale to warn you in time.

Sources

• ICH E6(R3) Good Clinical Practice