Claustrophobia and premature termination of magnetic resonance imaging examinations




To evaluate the incidence of MRI-related claustrophobia and prematurely terminated MRI (ptMRI) examinations due to claustrophobia in a large-scale cohort study.

Materials and Methods

The hospital's computerized radiology information system (RIS) was retrospectively analyzed for all 1.5-Tesla MRI examinations and reports during the year 2004. Data collected included demographic information, body part examined, known claustrophobia, and whether the examination was prematurely terminated. All information available on the MRI examinations and the patient-based data (i.e., excluding any additional examinations per patient) were analyzed.


A total of 5798 MRI reports of 4821 patients were evaluated. A total of 95 patients (1.97%) suffered from claustrophobia and 59 (1.22%) prematurely terminated the examination due to claustrophobia. The incidence of ptMRI was higher in women than men (no statistical significance). The majority of patients with ptMRI were between 20 and 80 years old. Patients undergoing head MRI showed the highest incidence of ptMRI and those undergoing extremity, breast, or pelvic MRI had the lowest. Prone compared to supine positioning results in the lowest ptMRI incidence (P < 0.05).


Claustrophobic reactions cause a relatively low incidence of ptMRI and are influenced by sex, body part examined, and positioning within the MR scanner. Sedation and prone positioning might help overcome these reactions. J. Magn. Reson. Imaging 2007. © 2007 Wiley-Liss, Inc.

SINCE ITS INTRODUCTION in the 1980s, MRI has become a popular diagnostic procedure. MRI evaluation usually involves the patient being inserted head-first into a relatively narrow tunnel and remaining motionless for up to and beyond one hour. Anxiety-related reactions during MRI have been extensively reported and researched, with a reported incidence of up to 37% (1–4). Claustrophobia is an anxiety disorder that involves the fear of enclosed or confined spaces (5). The reported incidence of premature termination, or failure of the MRI examination due to claustrophobic reactions ranges between 0.5% and 14.5%. The available data has mostly been obtained with relatively old MRI scanners and in small study cohorts (1, 2, 6–9).

The purpose of the current study was to evaluate the incidence of MRI-related claustrophobia and prematurely-terminated MRI examinations in a large-scale cohort study of patients undergoing 1.5-Tesla MRI.


Our university-based department of radiology uses a computerized radiology information system (RIS). For every MRI examination performed, the patient and examination data are manually stored in the data base. Additional information (i.e., if a patient misses the MRI appointment, prematurely terminates the examination, etc.) is also coded into the RIS.

An extensive computerized search of all MRI examinations and reports performed during the year 2004 was carried out. For each examination, data collected included demographic information of the patient (age, sex), body part examined, whether contrast medium or premedication was administered, known claustrophobia, and whether the examination was prematurely terminated. Only routine MRI examinations were included, while all study-based MRI examinations were excluded. Then, a manual search of all written MRI reports was performed to verify the computerized data and the information about the reason for premature termination and for premedication indications was added. Terminated MRI examinations for reasons other than known claustrophobia (N = 97) as well as reports with incomplete or unclear data (N = 14) were excluded (see summary in Table 1). Children under the age of five years (N = 30) were also excluded.

Table 1. Excluded Studies*
CauseNumber of examinationsIncidence (%)
  • *

    Causes of premature termination of MRI examinations other than claustrophobia in all 5939 examinations.

Pain and health condition480.81
Technical reasons50.08
Reaction to contrast agent40.07
Patient not metal-free30.05

Claustrophobia-related termination of the MRI examination was defined as severe anxiety (5) upon being positioned in the MRI tunnel with the patient requesting to be taken out while the attending radiologist could not reassure him or her. Patients were interviewed immediately after the incident to give the reason why they asked for the termination. Interview results were stored on file and in the reports. Though not verified by using specific psychological tests, the definition of claustrophobia is well understood by the attending personnel who are trained in regular team meetings and teaching sessions on the safety of MRI.

MRI units used were Magnetom Vision II, Magnetom SymphonyVision, and Magnetom Sonata (all Siemens Medical Solutions, Erlangen, Germany). Bore dimensions are summarized in Table 2.

Table 2. Magnet Dimensions*
MRI unitTunnel length (cm)Bore width (cm)Bore height (cm)
  • *

    Magnet bores of the MR scanners used in the study.

Magnetom Vision II2026058
Magnetom Symphony Vision1416060
Magnetom Sonata906060

A patient-based analysis was performed; i.e., for each patient only one examination was taken into account and any follow-up or other MRI examinations performed were excluded in order not to overestimate claustrophobia. Correlation was analyzed between prematurely terminated examinations and gender, age group, and body part examined.

Statistical Analysis

Descriptive statistical analysis was performed using SPSS for Windows version 14 (SPSS Inc., Chicago, IL, USA). The chi-squared test was used to identify statistical significance of the results. Statistical significance was defined as P < 0.05.


A total of 5939 MRI examinations were scheduled for the year 2004.

Out of the planned examinations, 5798 MRI reports of 4821 patients fulfilled the inclusion criteria and were evaluated.

The female-to-male ratio was 2585:2236 (54:46%); age range was five to 99 years, mean age was 49.9 years.

Patient-Based Analysis

A total of 46 patients (0.95%) reported known claustrophobia and required premedication with sedatives (intravenous midazolam at a dose of 2–5 mg) in order to perform the MRI examination, 10 of them prematurely terminated the examination despite the premedication. A total of 49 (1.02%) patients had newly observed claustrophobia requiring premature termination of the MRI examination. That is, the total number of patients with claustrophobia in our cohort was 95 (1.97%), and 59 (1.22%) patients prematurely terminated the examination due to claustrophobia.


A total of 38 males (0.79%) and 57 (1.18%) females had claustrophobia (no statistical significance). A total of 24 males (0.50%) and 35 females (0.73%) prematurely terminated the examination due to claustrophobia (no statistical significance).


Claustrophobia and prematurely-terminated MRI examinations due to claustrophobia occur in almost all age groups; however, only minor percentages were recorded below the age of 20 years and none over the age of 80 years (Fig. 1).

Figure 1.

Distribution of claustrophobia-related premature MRI termination by age group. Patient-based analysis.

Part Examined

The distribution of prematurely terminated examinations according to the part examined is summarized in Table 3. The highest incidence of claustrophobia-related examination termination is evident in head and neck MRI exams (1.73%), and lowest in breast and pelvis MRI exams (0.31% and 0.27%, respectively).

Table 3. Distribution of Claustrophobia-Related Premature Termination by Body Part Examined*
Body partNumber of patientsNumber of patients with claustrophobiaIncidence (%)
  • *

    Patient-based analysis.

Upper extremity5723.51
Lower extremity16910.59

A significant difference was found between examinations performed in the prone position (breast) and those performed in the supine position (all the rest) (P < 0.05)


Claustrophobic reaction is a well-known complication and disadvantage of MRI examinations, consuming both time and money. Thus, pharmacological sedation, hypnosis, and other relaxation techniques have been proposed to enable completion of the MR examination (7, 8, 10–12). However, only few studies tried to evaluate the incidence of MRI-related severe claustrophobic reactions and most of them evaluated it during a short period of time and on a relatively small cohort (1, 2, 6, 7). Only two studies investigated this issue in a large patient cohort (8, 9). To the best of our knowledge, our study is the only one reporting the incidence of claustrophobia in a large cohort on a patient level. It is obvious that patients suffering from claustrophobia would have the same symptoms at subsequent examinations. Thus analyzing the data according to examinations and not to patients would bias the results. In the current study, the total incidence of claustrophobia was found to be 1.97%. However, some patients were able to complete the examination under partial sedation, and an incidence of 1.22% was found for premature termination of the examination due to claustrophobia. A study published in 1990 reports an incidence of 1.5% out of 3000 examinations (8). Patients required sedation with intravenous diazepam due to severe panic attacks when undergoing 0.5-Tesla MRI. However, all patients were able to complete the examination following sedation. The second study, performed in Malaysia, does not report the total claustrophobia incidence but rather only the incidence of failed MRI examinations due to claustrophobic reactions (9). The reported incidence in that study is 0.54% out of 3324 MRI examinations.

The incidence of premature termination due to claustrophobia in all 5798 MRI examinations performed in our study (1.07%) as well as in the patient-based analysis (1.22%) is higher than reported in the other two large studies but still much lower than in earlier studies investigating smaller cohorts. Sarji et al (9) explain the lower incidence in their study by a relatively higher education level of their patients. Since we did not evaluate this parameter, we cannot verify this assumption.

More women had claustrophobia than men and more examinations were prematurely terminated in women compared to men. However, this difference was not statistically significant. Several studies reported a higher incidence of MRI-related claustrophobia and panic attacks in women (1, 7). However, this result is not always statistically significant. This tendency is also consistent with the higher prevalence of females suffering from anxiety disorders in the general population (13). In contrast, Sarji et al (9) reported a significantly higher incidence in men (67% males and 23% females) among the patients with prematurely terminated examinations. We cannot explain this difference, which may be due to cultural aspects.

The majority of patients suffering from claustrophobia and patients who prematurely terminated the examination ranged between ages of 20 and 80 years, with no claustrophobia apparent below 10 years and above 80 years of age (Fig. 1). No such analysis was performed by other authors. Sarji et al (9) reported that 78% of the patients with discontinued MRI examination were between ages of five and 45 years but do not elaborate on the rest of the patients and their proportion out of the total population studied. Our results correlate with the previously described tendency of patients older than 70 years to tolerate well MRI (14) and with the overall lower prevalence of anxiety disorders in patients older than 65 years (13).

The majority of claustrophobia-related premature MRI terminations occurred during head and neck examinations (Table 3). This is in concordance with other studies, and probably reflects the exacerbated confinement feeling of a patient within the head coil (7). Conversely, patients feel less confined when they undergo MRI of the legs, which explains the relatively lower incidence of premature terminations of MRI when the lower extremities are examined. Patients undergoing MRI of the distal extremities feel least confined because only the extremity is within the bore. This was indeed reported by McIsaac et al (2). In their study, patients positioned head-first within the MRI bore showed significantly higher anxiety reactions compared to patients positioned legs-first.

The high incidence of claustrophobia-related examination termination in MRI of the upper extremities must be considered artificial since the total number of exams (N = 57) included in the evaluation is rather small.

An interesting result is the very low premature termination incidence of breast and pelvis examinations (0.31% and 0.27%, accordingly).

All breast examinations are performed with the patient in the prone position. It was previously described that prone positioning of a patient might alleviate claustrophobic reactions (15). This could explain the low incidence of prematurely-terminated examinations in patients undergoing breast MRI. In patients undergoing pelvic MRI examinations the low incidence of prematurely-terminated examinations might be attributable to the relative distance of the patient's head from the surface coil.

Our study has some limitations; its retrospective nature is a major limitation. The total number of upper extremity MRI examinations is very low compared to the other body parts. This is why we regard the incidence of claustrophobia-related premature MRI termination in this region as artificially high and do not draw any conclusions from this number. Also, in the current study the radiologist present in the MRI suite defined claustrophobia according to the individual patient's behavior and reactions. It could be argued that some of the premature terminations of MRI were attributable to other types of anxiety reactions, like medical test anxiety, and not true claustrophobia. Usually, patients suffering from claustrophobia are well able to communicate their feelings and the attending personnel in our hospital are well trained to identify the symptoms of claustrophobia. In addition, the study objective was to evaluate the prevalence of claustrophobia-related premature terminations of MRI examinations and not the exact underlying nature of the claustrophobic reaction. If indeed some of the prematurely-terminated examinations resulted from other anxiety reactions, it may be assumed that the incidence of true claustrophobia is even lower than calculated in our analysis.

In conclusion, a large cohort, patient-based analysis of the incidence of MRI-related claustrophobia and the resulting premature termination of the examination is reported. Evidently, claustrophobic reactions still pose a problem and lead to premature termination of MRI examinations. Partial sedation or other anxiolytic methods as well as prone positioning might help to reduce this problem.