(Update: Statements from American Airlines and United Airlines have been added. StellaService removed American from its Twitter assessment after discovering it had tweeted questions to an inactive Twitter handle. The company also changed United's ranking to reflect the airline's assertions it had addressed questions via its inactive Continental handle.)

NEW YORK ( MainStreet) -- Airline passengers are accustomed to being stranded at the gate or on the tarmac, but Hurricane Irene left many fliers stranded on the phone and in the land of Twitter awaiting a response from the major carriers.

Many airlines had to cancel flights over the weekend, but not all of them handled Irene equally in delivering customer service to their passengers. StellaService, a company that rates customer service quality, ranked the 10 major U.S. airlines based on how they handled a series of calls to their customer services lines and requests for help on Twitter and found passengers on American Airlines ( AMR - Get Report) were stuck in the longest telephone holding pattern by far.

The airline had the longest wait time on its customer service line, keeping passengers on hold for an average of 1 hour, 32 minutes and 39 seconds. That was almost an hour longer than the next worst airline, Delta ( DAL), which had an average wait time of 33 minutes and 43 seconds.
A survey finds customer service at American Airlines foundered during Hurricane Irene, while U.S. Airways addressed customer queries quickly.

American also wasn't responding to Twitter messages, an area in which Delta actually excelled. The airline answered 100% of customer tweets in an average of 14 minutes.

But it was U.S. Airways ( LCC) that kept the phone lines flowing, answering calls in an average of just two minutes and 38 seconds. The airline was less successful with its social media support, though: It responded to only 16.7% of tweets. On the bright side, for the tweets U.S. Airways did answer, it took an average of only 24 seconds for customers to get a response.

"Kudos to U.S. Airways for keeping average hold times under three minutes the day before the storm hit," StellaService wrote in a company blog post. "It's obvious we have choices when it comes to choosing an airline, and while there are sometimes slight price differences that make us lean one way or another, at the end of the day it's all about the customer experience."

American said it disagreed with the survey results and believed the sample size was too small to paint an accurate picture of its customer service during the measured time.

"We handled more than 100,000 calls on Friday, and during the period in question our customers waited an average of 21 minutes, far less than alleged and in line with most of our peers," a representative said. "Of the 78 tweets directed to us from Thursday through Sunday, a significant number of which did not request action, we responded to 46 tweets either publicly or privately to assist customers, and we also sent four proactive tweets with travel information related to the storm."

StellaService contacted each airline an average of eight times between 9 a.m. ET to 6:30 p.m. ET on Friday. Twitter rankings were determined by the airlines' responses to a dozen tweets sent directly to them between midnight and noon on Friday.

According to the survey, the average call hold times of the 10 largest airlines during the hurricane were as follows:
  1. U.S. Airways: 00:02:38
  2. Southwest Airlines: 0:08:10
  3. Continental Airlines: 00:08:15
  4. United Airlines: 00:12:04
  5. Spirit Airlines: 00:24:07
  6. JetBlue: 00:24:17
  7. AirTran: 00:27:52
  8. Frontier Airlines: 00:29:54
  9. Delta: 00:33:43
  10. American Airlines: 01:32:39

In the social media sphere, JetBlue ( JBLU) also shone, answering 83.3% of tweets in an average of 11 seconds. Frontier Airlines answered 100% of its tweets, but took an average of 4 minutes and 4 seconds to do so. United ( UAL), like American, did not respond to tweets at all.

United said the survey results were misleading, as its customer service department did not respond to the tweets in question after that it had already answered the same ones Stella Service had sent to its inactive Continental handle from its active one.

"Had we not answered the questions they tweeted to our inactive @continental handle, we would have replied to the questions they tweeted to our active @United handle, just as we replied to more than 200 other customer inquiries on Twitter," a representative for the airline said.

StellaService initially said United and American did not respond to tweets at all. It has since changed United's Twitter response rate to a 58% from a 0% after discovering the questions had been posed to a Twitter handle that had been deactivated after United and Continental merged.

StellaService removed American from its Twitter rankings for similar reasons but said it stands by all other findings from the study.

You can find a complete list of the airlines' Twitter performance on StellaService's blog.

>To submit a news tip, email: tips@thestreet.com.

Follow TheStreet.com on Twitter and become a fan on Facebook.