Debate: Retrospective vs. concurrent data collection

Should data be gathered pre- or post-discharge?

By Patrice Spath, ART
Brown-Spath Associates
Forest Grove, OR

Over the past years, quality management professionals have debated the advantages of concurrent vs. retrospective data collection. Concurrent collection is an activity that occurs while the patient is hospitalized or undergoing active treatment, while retrospective collection occurs after the patient has left the facility. The choice between the two is not easily made. Should 100% of data be gathered while the patient is hospitalized or undergoing active treatment? Is retrospective data collection more efficient? Should some data elements be collected concurrently and others retrospectively?

In the past, many hospitals chose to integrate performance measurement data collection activities with the job of the utilization review or case management staff. This decision was made in hopes of minimizing staffing requirements, enhancing productivity, and ensuring timely intervention when utilization or quality problems were identified. The greatest benefit of concurrent data collection is the ability to initiate interventions when problems are identified. Concurrent interventional opportunities range from early identification of problems, which allows referrals to physician advisors and department chairmen, to improvements in documentation for utilization review and coding purposes.

Secondary benefits of concurrent review include:
• timeliness of feedback to medical staff departments;
• reduction of retrospective record review;
• ability to obtain a working DRG for billing purposes.

As hospitals gain more experience in the use of integrated and concurrent data collection systems, some are refining their original decision because of the shortcomings of concurrent data gathering. Many have decreased or eliminated concurrent data gathering in favor of retrospective collection of information.

First, many organizations have found that concurrent intervention is possible but not feasible because medical staff leaders fail to cooperate. Physicians often are reluctant to intervene during the episode of care. Information must be provided after the fact to medical staff committees; only then will some action be taken. Other drawbacks include an inability to access records easily while patients are in the hospital, a shortage of work space at the nursing units, and a lack of information available in the medical record while the patient is still hospitalized, such as incomplete test results and missing pathology reports. Reviewers can miss the total picture when reviews are done concurrently.

Concurrent reviews usually more costly

In addition, concurrent review of patient records often is more costly than retrospective review in terms of staff time. Charts must be handled more than once to obtain the same amount of data that could have been obtained in one pass-through of the chart after discharge. Most organizations find that concurrent review must be supplemented by retrospective review to ensure a complete data set for performance measurement purposes.

Is concurrent review a reasonable choice for collection of performance measurement data, or do its drawbacks outweigh its benefits? In general, if performance measurement data are not analyzed until after the patient is discharged, concurrent collection may prove to be more problematic than advantageous. Users should determine their commitment to immediate intervention opportunities. If you cannot get medical staff and administrative support for concurrent problem solving, the identification of quality problems while the patient is hospitalized may prove to be a fruitless and frustrating activity.

The general trend seems to be toward using a blended — that is, both concurrent and retrospective — data-gathering system. In those organizations where medical staff leaders, physician advisors, or risk managers are willing to intervene immediately when quality problems are discovered, it is worthwhile to identify untoward events concurrently.

Concurrent data collection also enhances infection control and case management referrals when necessary. If data gathering can be incorporated into the process of care — pathway variance reporting, routine patient care activities, and so on — then concurrent collection can be more cost-effective. However, some data elements are more efficiently gathered after the patient’s discharge. This is especially true of outcome data that usually require a completed medical record as the information source.

When possible, get direct caregivers involved in documenting the outcome information necessary for performance measurement activities. Shown on this page is a checklist that can be placed on the front of all postoperative patients’ charts when they return to the surgical clinic for their first follow-up visit. (See checklist, above.) The attending physician or clinic staff is responsible for completing the questions. Following the patient’s clinic visit, the form is sent to the hospital’s quality management department. This process encourages the clinic staff to be involved in identifying surgical site infections, reducing the need for retrospective chart review by quality staff.

Rather than using retrospective data collection only as a backup for missed or incomplete concurrent review, identify those data items that will be gathered concurrently and those that are collected retrospectively. Once medical staff have set quality goals or objectives and selected performance measures, the quality management department can define data elements necessary to create those measures. In addition, quality staff must identify the data elements necessary to support measures for the ORYX project.

Once all the necessary data elements are identified, determine the most appropriate source for each item. The hospital’s information management plan can be used by the quality management department to see where data might be available, thus reducing duplication. Next, select the best time to gather each data element.

If some of the data elements are to be gathered retrospectively, then the individuals who will be responsible for gathering the information must be identified. If your organization wishes to centralize data collection activities, then the quality management department may choose to retain both concurrent and retrospective data acquisition responsibilities. By limiting the function of data collection, staff performing the activity can become more proficient.

However, don’t overlook the expertise of the health information management staff. The professionals responsible for diagnosis and procedure coding have the skills necessary for gathering data elements from patient records. They are a logical choice for performance measurement data collection if sufficient coding staff are available. Coders are generally paid less per hour than case managers. Therefore, it may be more cost-effective to use coding staff for retrospective data collection rather than asking higher-paid case managers to spend their time collecting data concurrently. Save concurrent data gathering for those data elements that must be collected concurrently because interventions will take place during the delivery of care — untoward events or inappropriate care that requires immediate action. If data are needed only for retrospective analysis of performance patterns, this information may just as easily be collected retrospectively.