Accuracy of self-reported severity and level of spinal cord injury

Twenty-eight individuals participated in this study. Table 1 provides demographic information.

Table 1 Participant demographic information.Accuracy of self-report level of injuryRegion of injury

Participants’ self-reported ROI matched Clinical ROI in 93% (N = 26) of cases (χ2 = 23.713, p < 0.001; Γ = 1.000, p < 0.001). There were two exceptions: one participant whose Clinical ROI was cervical self-reported “thoracic,” and one whose Clinical ROI was thoracic self-reported “lumbar.”

Single neurologic level

When selecting Self-report LOI, only 25% selected the same level as their Clinical LOI, but 75% were accurate or within one spinal cord level (see Table 2). There was a strong, significant relationship between Clinical LOI and Self-report LOI (χ2 = 210.522, p = 0.002; rs = 0.922, p < 0.001). The average difference between Clinical LOI and Self-report LOI was −0.3571 with a standard deviation of 1.5448 and ranges from −5 to 2. The median was 0.00. The modes were 1, 0 and −1.00.

Table 2 Self-Report Level of Injury vs. Clinical Level of Injury and Initial Level of Injury.

When comparing Self-report LOI to Initial LOI, 61% selected the same level as their Initial LOI, and 79% were accurate or within one spinal cord level (see Table 2). The average difference between Initial LOI and Self-report LOI was −0.250 with a standard deviation of 1.174 and ranges from −3 to 2. The median and mode were 0.00.

Clinical LOI vs. initial LOI

Initial LOI and Clinical LOI were similar, with 28.6% being the same level and 53.6% with one level of difference (see Table 3). There was a strong, significant relationship between Initial LOI and Clinical LOI (χ2 = 256.265, p < 0.001; rs = 0.907, p < 0.001). The average difference between Initial LOI and Clinical LOI was −0.1071 with a standard deviation of 1.499 and ranges from −5 to 2. The median was 0.00; the mode was 1.

Table 3 Clinical Level of Injury vs. Initial Level of Injury.Accuracy of self-report AIS

When asked “Which of these is your ASIA grade?” 50% of participants responded, “I don’t know.” There was notable variation across AIS Classifications for percent of respondents choosing “I don’t know”: Clinical AIS A 33.3% (N = 2), Clinical AIS B 57.1% (N = 7), Clinical AIS C 14.3% (N = 1), and Clinical AIS D 87.5% (N = 7). There were no significant relationships between not knowing AIS grade and participant age (χ2 = 2.476, p = 0.780), gender (χ2 = 0.000, p = 1.000), household income (χ2 = 4.169, p = 0.654), time since injury (χ2 = 0.000, p = 1.000) or racial group (χ2 = 2.191, p = 0.139).

For the 50% who responded with an AIS grade, there was a strong, significant relationship between Clinical AIS and Self-report AIS (χ2 = 30.00, p < 0.001; Γ = 1.000, p < 0.001). In total, 71% (N = 10) self-reported the same AIS grade as their clinical exam including all those with injury classified as Clinical AIS A (100.0%, N = 4), Clinical AIS B (100.0%, N = 3) and Clinical AIS D (100.0%, N = 1). Of the participants who were Clinical AIS C, more than half self-reported as AIS B (66.7%, N = 4). The rest self-reported correctly as AIS C (33.3%, N = 2) (see Fig. 2). Thus, participants who reported AIS Classification correctly were 36% of the full sample.

Fig. 2: ASIA Impairment Scale—Self-Report vs. Clinical AIS/Initial AIS.figure 2

Number of respondents in each Self-Report AIS category who were clinically classified into each AIS group currently (Clinical AIS) or during their acute rehab (Initial AIS). **Does not include responses of “I don’t know” my AIS grade.

Eighty-six percent of participants who reported an AIS grade selected the grade given to them during their inpatient rehab admission. There was a strong, significant relationship between Initial AIS and Self-report AIS classification during the study (χ2 = 23.600, p = 0.001; Γ = 1.000, p < 0.001). For participants with Initial AIS A, 80% (N = 4) self-reported AIS A and 20% (N = 1) self-reported AIS B. All participants with Initial AIS B (100%, N = 6) self-identified as AIS B. Of participants with Initial AIS C, 66.7% (N = 2) self-reported AIS C and 33.3% (N = 1) as AIS D.

Clinical AIS vs. initial AIS

Eleven of 28 (39%) of participants have a current AIS classification that is different than their classification during inpatient admission. Only one of these changed to a more severe classification. There was a strong, significant relationship between Initial AIS and current Clinical AIS (χ2 = 34.746, p < 0.001; Γ = 0.877, p < 0.001). For participants with Initial AIS A, 66.7% (N = 6) are currently classified as AIS A, 11.1% (N = 1) are classified as AIS B, and 22.2% (N = 2) are classified as AIS C. For participants with Initial AIS B, 62.5% (N = 5) are still classified as AIS B and 37.5% (N = 3) are now classified as AIS C. Of participants with Initial AIS C, 14.3% (N = 1) are now classified as AIS B, 28.6% (N = 2) are still classified as AIS C, and 57.1% (N = 4) are now AIS D. Of participants with Initial AIS D, all (100.0%, N = 4) are still AIS D.

Accuracy of questionnaire/decision tree algorithm for determination of severity classification

Using the injury severity decision tree, 75% of participants (N = 21) were classified correctly relative to the Clinical AIS (see Fig. 3). Of Clinical AIS A participants (N = 6), 66.7% were classified correctly as complete and 33.3% were classified as sensory-incomplete. Of Clinical AIS B participants (N = 7), 85.7% were classified correctly as sensory-incomplete with 14.3% (N = 1) incorrectly classified as complete. Clinical AIS C participants (N = 8) were classified correctly as motor-incomplete 57.1% of the time with 14.3% (N = 1) incorrectly classified as complete and 28.6% (N = 2) incorrectly classified as sensory-incomplete. Clinical AIS D participants (N = 8) were classified correctly 87.5% of the time as motor-incomplete and incorrectly classified as sensory-incomplete 12.5% (N = 1) of the time. There was a strong, significant (χ2 = 23.747, p = 0.001; Γ = 0.872, p < 0.001) relationship between Decision Tree Severity classes and Clinical AIS grades. There were no significant relationships between accuracy of the injury severity decision tree and participants’ age (χ2 = 3.517, p = 0.621), gender (χ2 = 0.207, p = 0.649), race (χ2 = 0.081, p = 0.776), or income (χ2 = 6.435, p = 0.386).

Fig. 3: Clinical AIS vs. Decision Tree Severity.figure 3

Number of respondents in each Clinical AIS category who were categorized into each Decision Tree Severity class.

Accuracy of self-report DAP sensation

Participants correctly self-reported ability to feel DAP 86% of the time. There was a strong, significant relationship between self-reported and clinically-classified DAP sensation (χ2 = 11.221, p = 0.001; Γ = 0.939, p = 0.005). For participants without Clinical DAP, most self-reported “no” when asked if they could feel DAP (83.3%, N = 5) with only one incorrectly self-reporting “yes” (16.7%, N = 1). Inversely, for participants with Clinical DAP present, most self-reported “yes” (86.4%, N = 19), but a small proportion incorrectly reported “no” (13.6%, N = 3). Of note, 93% (N = 26) of participants were correct in Self-Report S4/5 sensation.

Accuracy of self-report VAC

Participants correctly self-reported ability to voluntarily contract their anal sphincter 82% of the time. There was a strong, significant relationship between Clinical VAC and Self-report VAC (χ2 = 11.873, p = 0.001; Γ = 0.938, p < 0.001). For participants without Clinical VAC, nearly all self-reported “no” (93.3%, N = 14) with only one self-reporting incorrectly as “yes” (6.7%, N = 1). For participants with Clinical VAC, over half (69.2%, N = 9) self-reported “yes” while 30.8% (N = 4) self-reported incorrectly as “no”. Thus, self-reported measures largely aligned with clinical measures for VAC, but participants without VAC were most likely to align their self-report to clinical measures.

Comparison of clinical S1, S4/5 and DAP sensation

In this sample, clinical S1 sensation and clinical S4/5 sensation were in agreement for 86% of participants. Clinical S1 sensation agreed with clinical DAP for 75% of participants and clinical S4/5 sensation agreed with clinical DAP for 89% of participants.

Accuracy of self-report of sensation at S1 (see Fig. 4)

Eighty-two percent of participants correctly reported ability to feel light touch (LT) at the heel/S1 dermatome. There was a strong, significant relationship between clinical and self-reported LT S1 (χ2 = 12.253, p < 0.001; Γ = 0.941, p < 0.001). Nearly all participants who were clinically classified as “no” for LT S1 also self-reported that they did not have “the ability to feel a LT with cotton wool on the skin on either of their heels” (92.3%, N = 12). Only one participant who was clinically classified as “no,” self-reported incorrectly that he had LT sensation on his heels (7.7%, N = 1). Most participants who were clinically classified as “yes,” having LT sensation present, also self-reported “yes” (73.3%, N = 11) with a few incorrectly self-reporting “no” (26.7%, N = 4).

Fig. 4: S1 sensory and motor sparing—Self-Report vs. Clinical Exam.figure 4

Number of respondents self-reporting presence (YES) or absence (NO) of each modality who were clinically found to have (YES) or not have (NO) sparing of that modality. S1 LT light touch sensation, S1 PP sharp/dull discrimination, S1 Motor motor function (muscle strength ≥1/5).

Eighty-six percent of self-report responses were correct for sharp/dull discrimination at the S1 dermatome (PPS1). There was a strong, significant relationship between clinical PPS1 and participant-reported “ability to tell the difference between the sharp and blunt ends of a safety pin on the skin of either heel” (χ2 = 13.741, p < 0.001; Γ = 0.942, p < 0.001). Of participants who were clinically classified as “no” for PPS1, 88.2% (N = 15) reported “no” for ability to tell the difference between sharp and dull, with 11.8% (N = 2) incorrectly reporting “yes”. Of participants who were clinically classified as “yes,” 81.8% also self-reported “yes” (81.8%, N = 9) with only two participants, self-reporting “no” (18.2%).

Accuracy of self-report of motor sparing at S1

Ninety-six percent of self-report responses were correct for motor sparing at S1. There was a strong, significant relationship between clinical motor function at S1 (plantar flexion >0) and participant-reported “ability to move either ankle as if pushing down a gas pedal” (χ2 = 24.040, p < 0.001; Γ = 1.000, p < 0.001). All participants who were clinically classified as “no” for Motor S1 also self-reported “no” for Motor S1 (100.0%, N = 17). Most participants who were clinically classified as “yes” also self-reported “yes” (90.9%, N = 10) with only one participant who was clinically classified as “yes” incorrectly self-reporting “no” (9.1%, N = 1).

Comparison of S1 sensory/motor function and DAP/VAC

In this sample, self-report of S1 sensation agreed with Clinical DAP for 79% of participants. Presence of clinical Motor S1 agreed with Clinical VAC for 86% of participants. Self-report of motor function at S1 matches Clinical VAC for 89% of participants.

留言 (0)

沒有登入
gif