“Backlog,” an indication of work to come, is not an official accounting term. And yet, due to the uneven, contract-based nature of the business, publicly traded CROs like to include it high up in their quarterly reports.

Using backlog allows CROs to put a positive spin on what otherwise might be less-than-stellar earnings; while they may not have had the healthiest quarter—or year—financially, they can showcase all of the work that has been promised over the next few quarters in millions, or billions, of dollars.

In the last few years, with the advent and popularity of strategic-partnership agreements between CROs and large pharmaceutical companies, backlog numbers are looking markedly robust. After all, these deals in most cases promise CROs many large trials over multiple years.

Consider Parexel, which, along with Icon, landed a huge strategic-partnership agreement with Pfizer this summer. On Aug. 9, the company reported fiscal year-end backlog of $3.44 billion, up 28.5% year-over-year. That number included gross new business wins in the last quarter of $675.1 million. PPD’s second quarter results, announced July 26, included a massive $3.7 billion in backlog, representing growth of 14% over the year-ago quarter. PPD inked a large strategic partnership agreement with GlaxoSmithKline in September 2010, as did Parexel.

There’s no doubt backlog is swelling among the elite CROs winning such large deals. But some say huge backlog can come back to bite a company, especially if the work is not due to start for some time. And that’s the nature of much strategic-partnership work, said Michael Martorelli, director at investment bank Fairmount Partners and consultant to the CRO industry.

“These healthy backlogs can translate into high cancellations,” Martorelli said.

That’s because this industry is rife with cancellations, so work promised in earnest today could easily evaporate tomorrow. Big work. For example, PPD said in its most recent quarterly earnings report it had to reduce backlog by a whopping $43.2 million due to the cancellation of just one study.

Martorelli added that such massive backlog numbers also increase already existing confusion about whether backlog refers to work that’s been promised via contract, or simply work that has been discussed with possibly a hand shake. Each company seems to have its own interpretation.

Martorelli points to Covance, which, to avoid confusion about the exact nature of what’s in its backlog, offers two numbers. A Covance spokesperson explained the company has a subset of regular backlog called “Contractual Minimum Volume Commitment,” or CMV, which includes only work that has been committed to via contract and is being paid for at regular intervals, akin to a retainer. Some of the work in this category is from strategic agreements and some is not, she said. Regular backlog, which is “no commitment involved,” includes work from strategic agreements as well as work that is not. For the second quarter, Covance said its overall backlog had gone up 29% to $6.25 billion—$2.42 billion of which was CMV and $3.83 billion of which was not.

“In a way this confuses the issue because you have two different backlogs,” said Martorelli, “but at the same time, it’s also much more illuminating.”

Another consequence of bigger backlog is reducing the speed at which upcoming work turns into revenue, said William Blair analyst John Kreger.

“In general, strategic partnerships have slowed the conversion of backlog into revenues by awarding studies earlier,” he said, explaining that in the 1990s, work was often awarded to CROs “at the eleventh hour,” and even five years ago CROs had about three to six months’ notice before a study needed to start. “Now, they may get awards 12 months in advance.” This leads to a lag between the addition of the work to the backlog and the addition of the payment from the work to the bottom line.

Also, massive backlog can be off-putting to other potential suitors. Said Martorelli, “Sponsors may say, ‘Your backlog is huge. I probably won’t get the attention from you that I used to. We’ll give our work to a mid-tier CRO that’s not so busy.’”

This, in turn, he said, could start to produce more work for the mid-tier CROs in the industry.

Suz Redfearn