Concept: Computed Tomography (CT) Imaging

What is a CT Scan?

This concept describes how MCHP has identified and defined computed tomography (CT) imaging in its research using the Manitoba Population Research Data Repository.

A "CT" or "CAT" scan is the term used to describe a radiologic test known as "computerized tomography." The scanner generates images of thin slices of any part of the body. CT scans are a key diagnostic procedure for examination of internal body structures. All urban hospitals are equipped with CT scanners, with HSC and St. Boniface each having two units.

This concept also provides a link to relevant SAS code. Please see the SAS code and formats section containing the following link to: CT Scan Rates SAS Code - (internal access only).

Method of Identifying CT Scans

Records for CT usage only became comprehensive in 1998/99, so years prior to this cannot be analyzed (outpatients at community hospitals were not recorded before that time).

In the Medical Services/Physician Claims data, CT scans are coded using tariff codes. The following codes were developed for Manitoba medical and hospital claims beginning in 1998/99. The range of codes is consistent over the years since then, but should always be verified prior to use. Only individuals that are registered with Manitoba Health and had a valid Manitoba hospital number were used in the analysis.

CT scans were identified from the Medical Services/Physician Claims data with any of the following tariff codes:


• 7112 - 7115 and 7221 - 7230 (inclusive)

Note about CT Scans for Inpatients, Outpatients and in the Emergency Room


In Frohlich et al. (2001), they investigated the type of patients receiving CT scans. Unfortunately, in the Hospital Discharge Abstracts data, the OPD/Hospital fields should not be used alone to indicate if the patient was an inpatient, outpatient or emergency. We overcame this problem by identifying everyone who received a CT scan, then checked hospital discharge abstracts to see if the scan date fell within an admission-separation date range (for inpatients). If so the scan was done during an inpatient stay and we called them inpatients.

The CT Scan data from Medical Services was sorted and merged to Hospital Discharge Abstract data for the years 1998/99 and 1997/98 by (scrambled) PHIN and hospital; only records found in both files were kept. The service date on the physician claim was compared to the admission/separation date range on the hospital discharge abstract to identify scans that were done within a hospital stay. The MCHP variable TRANSACT was used to identify inpatients and outpatients since the OPD variable in the Medical Services data was found to be not reliable enough for this analysis.

These data were merged back to the Medical Services data using (scrambled) PHIN, hospital number and service date as identifier. Inpatients were identified using TRANSACT= 1; all other records were coded as Outpatient.

Further Information

For more information, see:


• section 10.8 Computed Tomography (CT) Scans in The Manitoba RHA Indicators Atlas by Martens et al. (2003)
  
• section 6.6 Computed Tomography (CT) Scans in Sex Differences in Health Status, Health Care Use, and Quality of Care by Fransoo et al. (2005)
  
• section 8.8 Computed Tomography (CT) Scans in The 2013 RHA Indicators Atlas by Fransoo et al. (2013)
  
• section 8.7 Rates of Computed Tomography (CT) Scans in The 2019 RHA Indicators Atlas by Fransoo et al. (2019)

Cautions / Limitations

• To count person-visits, only one scan per day is counted, as there could be multiple body parts scanned, each with their own claim.
  
• Individual-level information regarding CT scans performed in rural hospitals are not always recorded. Therefore, CT scan rates reported may under-estimate the "true" CT scan rates to an unknown degree.
  
• In Fransoo et al. (2013, 2019), the analysis of CT scans was limited to Manitoba residents aged 20 years and older.
  
• The relationships with premature mortality rates are mixed: there is a significant relationship at the neighbourhood cluster level, but a relationship that falls short of significance at the community level.