Results
During this cohort period, INCA Pediatric Oncology registered 1,625
patients, from these, 917 patients had the diagnosis of cancer, 157 were
primary bone cancer who had been included.
Fifteen patients (9.6%) were loss of some data, but these patients were
analyzed: 9 (60%) with diagnosis of OS and 6 (40%) ES; median age 12
years old (IQR 8 - 14) and 4 (26.7%) died. Comparisons were made
between loss of data patients and the cohort sample according to age
(Mann-Whitney Test; p = 0.42), diagnosis (Fisher Exact Test; p = 0.30)
and survival (Fisher Exact Test; p = 0.24).
The cohort study comprised 142 patients, with median age 13 years old
(IQR 9-14 years old), 43.6% male sex, being 99 OS (69.7%) and 43 ES
(30.3%). Median time for initial treatment after registration was 22.5
days (IQR 16.0 - 34.0 days) and 72 (50.7%) patients had metastatic
disease at presentation, 46/99 OS (46.46%) and 26/43 ES (60.46%).
Patients characteristics at registration are in Table 1.
We examined pain characteristics and previous treatment at registration.
Seventy-six patients (53.5%) had pain evaluated by a method at that
moment and 99 (69.71%) were in previous use of pain treatment.
At registration, pain treatment was modified in 60 patients (42.25%);
pain treatment at that moment evaluation were prescribed to 103 patients
(72.53%) in the following frequency: mild analgesics in 42 (29.57%),
weak opioids in 37 (26.05%) and strong opioids in 17 (11.97%), strong
opioid plus mild analgesics in 1 (0.70%), strong opioid plus mild
analgesics in 6 (4.22%), anticonvulsants in 11 (7.75%),
antidepressants in 21 (14.79%) and steroids in 3 (2.11%). Three
patients with no pain had their pain treatment discontinued. It was
initially observed a reduction in the use of mild analgesics and an
increase in the use of weak and strong opioids.
Three months after registration, patients were evaluated again. Eight
patients (5.63%) had disease progression. Seventy-one patients (50%)
had their pain treatment changed. Pain treatment at that moment was: 21
(14.78%) mild analgesics, 29 (20.42%) mild opioids, 55 (38.73%)
strong opioids, 4 (2.81%) mild and strong opioids, 2 (1.40%) mild
analgesics and weak opioid, 36 (25.35%) anticonvulsants, 60 (42.25%)
antidepressants and 4 (2.81%) steroids.
At last evaluation, all 142 patients had received antineoplastic
treatment, 105 (73.94%) were submitted to surgery, 31 (21.83%) to
radiotherapy and 62 (43.66%) to chemotherapy. Sixty-five patients
(45.77%) had disease in progression. Seventy-three patients (51.41%)
had their pain treatment changed. Pain treatment at that last evaluation
was: 19 (13.38%) mild analgesics, 21 (14.78%) mild opioids, 63
(44.36%) only strong opioids, 3 (2.11%) weak and strong opioid (Figure
1).
The comparison of pain at registration, three months later and at last
evaluation showed along the time a more frequent pain assessment and
absence of patients with excruciating pain (Table 2).
A hundred and twenty-two patients had pain classified at last
evaluation. Twenty patients who did not have pain assessment by scale,
reported no pain at last evaluation and were classified as no pain.
According to this, data was imputed based on pain complaint at last
evaluation (self-assessment). Analysis between those who survived or
died, revealed no difference of pain status between groups (p = 0.68).
Fifty-six
patients evolved to death (39.4%), it was observed a 25% (14/56)
frequency of pain complaint among those end-of-life patients and a 22%
(19/86) among the others at last evaluation (Table 3).
Poisson regression to estimate CPR of pain (yes/no) at last evaluation
revealed no difference of pain status between those end-of-life or not
(p=0.68), as can be seen on Table 4. All variables with CPR p value
<0.20 were candidates to enter a multivariate model to predict
adjusted prevalence ratios (APR).
A multivariate model to estimate APR included disease progression, use
of strong opioids and steroids at last evaluation. This model was
adjusted for age (Table 5).