- Email: [email protected]
Pain progression, intensity and outcomes following tonsillectomy
Pain 75 (1998) 37–45
Pain progression, intensity and outcomes following tonsillectomy Fay F. Warnock*, Janice Lander 3-132 Clinical Sciences Building, University of Alberta, Edmonton, Alberta, T6G 2G3, Canada Received 4 August 1997; received in revised form 8 October 1997; accepted 7 November 1997
Abstract The objective of this study was to assess outcomes of pediatric day surgery tonsillectomy. A total of 129 children, aged 5–16 years, and their parents were recruited from three urban hospitals which provided pediatric day surgery. Children reported pain on a visual analogue scale (VAS) in day surgery and then daily at home for 7 days. Parents reported outcomes of surgery, including fluid intake, nausea, vomiting and sleep disturbances. They also recorded analgesic administration. Three main results related to extent and duration of pain, quality of management of pain, and effect of pain on utilization of health services. Tonsillectomy caused considerable pain which lasted more than 7 days. Pain followed a trajectory of intense or moderately intense pain for the first 3 days followed by a gradual decline over the next 4 days. In general, post-tonsillectomy pain was poorly managed by health professionals and parents. An unexpected observation was that children who had a bupivacaine infiltration of the tonsil fossa during surgery had significantly more pain in the evening of surgery than children who did not have an infiltration. The increase in postoperative pain experienced by those who had the infiltration was attributed to quality of pain management. Children with persistent pain (those who did not follow the typical trajectory) were likely to be taken to a medical practitioner. One-third of the sample made unscheduled visits to practitioners with most occurring from Day 4 to Day 7 of the follow-up. 1998 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Tonsillectomy; Postoperative pain; Day surgery; School-aged child; Adolescents
1. Introduction Researchers have been aware for some time that children’s postoperative pain is managed poorly (Beyer et al., 1983; Mather and Mackie, 1983). Perhaps one explanation relates to a lack of information about prevalence and patterns of acute postoperative pain caused by surgical procedures, including tonsillectomy. Tonsillectomy is one of the most common of pediatric surgeries. In several studies that have been conducted, parents or staff have rated children’s postoperative pain (Gedaly-Duff and Ziebarth, 1994; Finley et al., 1996; Gabalski et al., 1996). According to parents, children’s post-tonsillectomy pain is moderately high on the first 2 days following surgery (Finley et al., 1996). Parent ratings, however, are not always in agreement with the child’s rating (see for example, Doherty et al., 1993; Benestad et al.,
1996). Despite their assessment of postoperative pain, parents under-administered analgesics (Gedaly-Duff and Ziebarth, 1994; Finley et al., 1996). Parents’ fears about tolerance, addiction and drug side-effects and their lack of knowledge about pain management are barriers to effective administration of analgesics (Gedaly-Duff and Ziebarth, 1994; Finley et al., 1996; Forward et al., 1996). Since many tonsillectomies are performed in day surgery, postoperative care including pain-management is being provided by parents in the home. To determine intensity and progression of pain as well as pain outcomes, we undertook a 7-day follow-up of children who had tonsillectomy in day surgery.
2. Methods 2.1. Subjects and setting
* Corresponding author. Tel.: +1 403 4312051; fax: +1 403 4922551; e-mail: [email protected]
This study was conducted in all hospitals in a mid-west
0304-3959/98/$19.00 1998 International Association for the Study of Pain. Published by Elsevier Science B.V. PII S0304-3959 (97 )0 0202-9
38
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Canadian city which provide pediatric day surgery for tonsillectomy for the area (n = 3 hospitals). The region is multi-cultural. Most of the children having surgery live in the city or within 40 km of it. Others come from outside of the city, some traveling as much as 800 km for surgery. All children, 5–16 years, having tonsillectomy day surgery were recruited during the 13-week sampling phase of the study. An additional inclusion criterion was that children be able to use a visual analogue scale. Surgical procedures performed on the study sample consisted of tonsillectomy alone or tonsillectomy with adenoidectomy and/or myringotomy. All procedures were performed by otolaryngologists or general surgeons and under general anesthesia. In addition, tonsil fossa of some children were infiltrated with bupivacaine HCl 0.25% or 0.50% with or without epinephrine during surgery. 2.2. Instruments A 100 mm visual analogue scale (VAS) was used for assessing pain intensity. The VAS is a valid, reliable and easily understood tool which has been used successfully to obtain self-reported pain from children ≥5 years of age (McGrath et al., 1985). The VAS was arranged as a 100 mm vertical line with anchors at either end (No Pain and Worst Pain Possible). Children’s comprehension and proficiency at using the VAS were assessed with an approach utilized in other research (Lander et al., 1992). Instructions about use of the VAS were provided to children who were then required to correctly rank sketched faces that illustrated pain. No child was excluded from this study because of inability to use the VAS. 2.3. Procedure The study procedure was divided into three phases. These included: pre-operative period, postoperative period while the child was in day surgery, and postoperative period at home. 2.3.1. Pre-operative phase Shortly after their arrival at the day surgery, the study was described to parents of children having tonsillectomy. If the parent agreed to participate, the child was recruited and instructed about the use of the VAS. The child’s ability to use the tool was then assessed. Parents provided information about the child’s general health status and history of previous hospitalizations. They were instructed about use of a medication diary. An envelope which contained VAS scales and a medication diary was provided for use at home. 2.3.2. Postoperative day surgery phase After tonsillectomy, children recovered in the day surgery for a period of time. Two to three hours after surgery,
children were asked to rate their current pain using the VAS. At the end of this phase, the research assistant recorded information from the hospital chart about the surgical procedure and prescribed analgesics. 2.3.3. Postoperative home phase Beginning with the day of surgery (Day 1) and every evening for 6 additional days (Days 2–7), parents were contacted by telephone in the early evening. On Day 1 parents reported for the period of time from when they returned home until the interview. For Days 2–7, they reported for the 24 h period of time since the previous interview. During each phone call, which took approximately 5– 10 min, parents were asked if the child had or had not experienced any of the following: pain; nausea and vomiting; postoperative bleeding; and uninterrupted night time sleep. They were also asked for the amount of fluid taken by the child. At the end of the interview, they were given the opportunity to describe any other concerns related to the child’s progress and care. After the parent was interviewed each evening, the research assistant asked to have the child brought to the telephone. The child was re-instructed about use of the VAS and asked to report current pain by making a pencil mark across the vertical line. A new VAS was used for each day’s pain report and the child was instructed to place the completed and dated VAS in the storage packet that had been provided. At the end of the 7-day follow-up period, parents were asked to return the VAS scales and medication log in the stamped and addressed envelope that had been provided. 2.3.4. Data preparation All data were coded and examined for completeness and accuracy. Distributions were inspected and descriptive statistics obtained for all variables. VAS pain ratings were computed by measuring the distance along the line to the point where the subject’s mark intersected the line. Parent reports of outcomes such as pain, nausea, vomiting, sleeping and bleeding, as reported in the telephone interview, were coded as dichotomous variables. Fluid intake was used as an indicator of risk of dehydration. Since others (Carabott et al., 1992) have set the minimum acceptable level of postoperative intake at 500 ml over 8 h, we coded an intake of ,30 ml/h as poor. Frequencies of all parent-reported outcomes were computed. A variable referred to as ‘percentage of daily dose administered’ was computed as follows: the total milligrams of analgesic administered was divided by the maximum amount of analgesic that could be administered, and the result was then multiplied by 100. This variable was computed separately for Days 2–7 for 108 cases (no dose had been specified for the other four cases). Day 1 was not included since the period of time that parents could have administered analgesics was variable depending on the time of surgery.
39
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
3. Results 3.1. Subject recruitment A total of 147 children, aged 5–16 years, had tonsillectomy during the recruitment interval. All were approached for study participation with the exception of five children who were taken to the operating room earlier than anticipated, before the arrival of the research assistant. Of those 142 approached, 10 did not participate for the following reasons: four children declined although their parents had consented and six parents declined because they lacked time or were reluctant to have their child involved in research. Therefore, 132 children were entered into the study. Two children were lost to follow-up when their families could not be reached by telephone (thus, n = 130). The number of cases available for analysis depended on the type of data being analyzed. Six children were admitted to hospital at some point during follow-up thus reducing the sample size for child-reported pain data and parent telephone interview data. One family did not complete any interviews as the child was admitted directly from day surgery. Several telephone interviews were conducted with four parents before or after the child was admitted. For one parent and child, all interviews and pain assessments were conducted as admission occurred on the last day of follow-up (this case was included in data analysis). The sample was reduced to n = 125 for pain data but n = 129 for telephone interview data. The VAS scales and drug administration logs were not returned or went missing in the mail for 13 subjects who completed all telephone interviews (n = 112). Therefore, analysis of the interview data was conducted with data from 129 cases whereas analysis of pain data was conducted with 112 cases. 3.2. Subject characteristics The children of parents who were interviewed (n = 129) had a mean age of 9.3 years (SD = 3.5). There were 64 girls and 65 boys. Approximately one third of these children had not been hospitalized previously and most were having their first surgery. Further details about characteristics of the sample are reported in Table 1. The rate of surgical procedures performed were: tonsillectomy with adenoidectomy (n = 81, 63%), tonsillectomy alone (n = 39, 30%), tonsillectomy with myringotomy and adenoidectomy (n = 9, 7%). No further analyses of surgical procedures were done because of insufficient numbers in two categories. 3.3. Trajectory of postoperative pain During the postoperative day surgery phase, the mean VAS score was 66.6 (SD = 32.7, n = 120). Nine young children (mean age 5.7 years) refused to rate pain while recovering in day surgery.
A repeated measures analysis of variance was used to assess changes in postoperative pain as reported by the children on the telephone for Day 1 to Day 7 (F = 70.7, P , 0.0001, n = 112). There was a significant drop in pain from Day 1 to Day 7. Mean daily ratings of postoperative pain intensity ranged from 64.8 (Day 1: SD = 29.2) to 26.8 (Day 7: SD = 25.0). Fig. 1 illustrates daily mean VAS scores. Frequencies of children reporting VAS pain ratings of ≥30 mm were determined for each of eight pain measures obtained. The frequencies were: Day Surgery, 97 (80.8%); Day 1, 99 (88.4%); Day 2, 84 (75.0%); Day 3, 69 (61.6%); Day 4, 51 (45.5%); Day 5, 51 (45.5%); Day 6, 47 (42.0%); and Day 7, 34 (30.4%). 3.4. Prescribed and administered analgesics Analgesics prescribed for postoperative pain were obtained from subject charts (n = 129). The drugs prescribed were: acetaminophen plain, n = 77 (60%); acetaminophen with codeine, n = 34 (26%); and codeine, n = 18 (14%). Percentage of daily doses administered were computed for analgesics as described previously. Correlations between daily child-reported pain and percentages of daily dose administered were near zero and non-significant except for Day 7 (r = 0.22, P = 0.02). By contrast, the correlations among percentage of daily dose administered were moderate to strong positive and significant. For example, the percentages administered in the first 2 days were correlated with the percentage administered on subsequent days (Day 2 with Day 3–7 correlations ranged from 0.37 to 0.66; and Day 3 with Day 4–7 correlations ranged from 0.59 to 0.79). Table 1 Subject characteristics (n = 129) Participants completing telephone interviews (%) Gender Boys Girls Past hospitalizations None Once More than once Previous surgeries None Once More than once General health status Excellent Very good Good Poor Residence Urban Rural
65 (50.4) 64 (49.6) 52 (40.3) 42 (32.6) 35 (27.1) 86 (66.7) 32 (24.8) 11 (8.5) 25 61 28 15
(19.4) (47.3) (21.7) (11.6)
101 (78.3) 28 (21.7)
40
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Receiving a low percentage of daily dose on Day 2 or 3 was therefore associated with receiving a low dose on subsequent days. Few children received the full dose each day, including on the days immediately following surgery. In fact on the day following surgery, half of the children received less than 50% of the analgesics that they could have received and only nine (8.3%) received the full dose. Several children received more than prescribed including one 5 year old who was given almost twice the prescribed and recommended dose of acetaminophen on 5 days of the follow-up. Night-time analgesic administration patterns were also examined for Day 1 to Day 7. On 4 of 7 nights, including the day of surgery, only 25% of the sample received analgesics at night-time. On the second day, about half of the sample were given analgesics during the night and about one-third during the third and fifth nights. Bupivacaine HCl was infiltrated at the tonsil fossa for 35 children (27%) who also had a general anesthetic. Their pain during follow-up was compared with those who did not receive infiltration. We were interested in examining overall differences in the two groups occurring over follow-up days (group by time interaction). Age, however, was correlated with child-reported pain (r = 0.20) and those who received tonsil fossa local infiltration were significantly older (mean 11.2 years, SD = 3.6) than those who did not (mean 8.8 years, SD = 3.3) (F = 11.0, P = 0.001). Thus, age was related to both grouping and outcome variables posing difficulties for data interpretation. We therefore matched each child who received a local infiltration with a child of the same age who did not receive a local infiltration. This strategy created a reduced sample of two
age-matched groups of n = 27 each, thereby controlling age. Equivalence of the matched groups was confirmed for gender, age and child’s weight (with analysis of variance or chi-square). Repeated measures analysis of variance indicated significant differences in pain for the two infiltration groups depending on follow-up time (F = 2.2, P = 0.04). Fig. 2 illustrates child-reported pain for the two groups over the follow-up period. Those who did not have the infiltration had a progressive decrease in pain over the follow-up period whereas the pain of those having the bupivacaine infiltration rose during the evening of surgery before progressively declining over the next 5 days. The infiltration group had significantly less pain in day surgery and more pain during the evening compared to the no-infiltration group (Day Surgery, infiltration group 64.8, no-infiltration group 80.3; Evening of Surgery, infiltration group 75.9, no-infiltration group 67.2). Percentage of analgesics were compared from Day 2 to Day 7 for the two age-matched infiltration groups (with repeated measures analysis of variance). There was a significant interaction between infiltration group and time (F = 6.0, P , 0.001). The no-infiltration group received significantly more analgesics on Days 2 to 4 compared to the infiltration group (Fig. 3). For example, on Day 2 the noinfiltration group received 63.5% of prescribed analgesics compared to 33.8% for the infiltration group. The age-matched groups were compared and significant differences in type of prescribed drug found (chisquare = 8.0, P = 0.02). The no-infiltration group was prescribed more acetaminophen plain and the infiltration group more codeine and more acetaminophen with codeine (Table 2).
Fig. 1. Mean pain ratings over 7 days (n = 112) 1 (DS) refers to VAS in day surgery on Day 1, 1 (Home) refers to VAS at home on Day 1.
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
41
Fig. 2. Mean pain for Infiltration and No-Infiltration groups. Groups matched by age (n = 27). 1 (DS) refers to VAS in day surgery on Day 1, 1 (Home) refers to VAS at home on Day 1.
3.5. Common postoperative occurrences As mentioned, six parents reported postoperative complications requiring admission to hospital. Two children were admitted directly from day surgery due to postoperative hemorrhage and three other children were hospitalized on the second postoperative day, two for ear and throat infection and one for throat pain and infection. Another was hospitalized on the seventh postoperative day for dehydra-
tion and hypoglycemia secondary to refusal to swallow fluids and food. Parents also reported occurrences of any medical consultations that were not part of pre-arranged postoperative care. Forty-nine (38.3%) parents reported having to make unscheduled medical visits or phone calls related to the child’s recovery from surgery. Forty required a single unscheduled consultation during the 7-day follow-up whereas the other nine required multiple consultations
Fig. 3. Mean percent of analgesics for Infiltration and No-Infiltration groups. Groups matched by age (n = 25)
42
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Table 2
Table 3
Frequencies of analgesics prescribed for age-matched infiltration and noinfiltration groups
Type and frequency of unscheduled medical consultations.
Acetaminophen plain (%)
Acetaminophen with codeine (%)
No-infiltration Infiltration
16 (59.3) 6 (22.2)
8 (29.6) 13 (48.1)
Total
22
21
Codeine (%) 3 (11.1) 8 (29.6)
Frequency of consultation Family physician Surgeon Walk-in clinic Phoned day surgery Emergency
37 7 6 6 5
Total
61
11
(two consultations, n = 6; three consultations, n = 3). Therefore, a total of 61 consultations were obtained. The majority (74%) of consultations were obtained on Days 4–7 (with 66% taking place on Days 5–7). Table 3 describes type of consultations made. Those who did and did not obtain unscheduled medical consultations were compared. The two groups were not significantly different for the following: (i) child’s health status, gender, history of previous hospitalizations or surgeries, whether or not parents had previous experience with a child in day surgery, whether they resided in an urban or rural area, and type of prescribed analgesic (using chi-square analysis); and (ii) parent’s expressed worry about caring for the child at home on a scale of 1 to 10 (using analysis of variance). A significant repeated measures analysis of covariance was obtained when consultation and no-consultation groups were compared for child-reported pain on Days 1–7 (child’s age was the covariate). The progression of pain for the no-
consultation group followed the trajectory plotted for the total group (as noted in Fig. 1). For the consultation group, pain declined initially, leveled off on Day 4 and then increased beginning on Day 5 (Fig. 4; F = 4.0, P = 0.001). Children of parents who required unscheduled consultations had significantly more pain, twice as much on average, on Days 5, 6 and 7. The two groups of parents who did or did not seek unscheduled consultations were compared also for percentage of drugs administered on Days 2–7 with repeated measures analysis of covariance (child’s age as covariate). There were overall significant differences in percent of analgesics administered for the two groups of parents (F = 5.5, P = 0.02). Mean overall percent administered for the no-consultation was 27.5% and 51.2% for the consultation group. As well, there was a significant interaction between consultation group and follow-up day (Fig. 5; F = 7.2, P , 0.001). On Days 5, 6 and 7, parents who
Fig. 4. Mean daily pain for groups obtaining or not obtaining medical consultation.
43
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Fig. 5. Mean percentage of drugs administered for groups obtaining or not obtaining medical consultation.
sought a medical consultation gave a significantly greater percentage of analgesics (52.8%, 53.5% and 54.0%, respectively) compared to parents who did not seek consultation (18.2%, 14.8% and 11.6%, respectively). Parents reported various outcomes during the telephone interview. Since parents could report multiple outcomes on any day and repeat the outcome from day to day, total frequencies for outcomes are reported in Table 4. Of note are the findings of inadequate fluid intake (which were reported during the first few days), and the occurrence of ear and jaw pain (which were reported increasingly in the latter days of the follow-up period).
beginning a decline again. The change in the downward trend of pain at Day 4 coincided with reports of onset of ear and jaw pain. Some clinicians have suggested that surgical trauma and postoperative inflammation affects coordination of nasopharyngeal and eustachian tube muscles (Holt et al., 1981). This may be one explanation for occurrence of ear and jaw pain. An interruption in the downward trajectory of throat pain perhaps stems from the return to a normal diet (with concomitant increase in swallowing) as the child convalesces. On Day 7, when pain was at its lowest, almost one-third of children rated pain as 30 mm or more on the VAS. This was the criterion used by Finley et al. (1996) for clinically significant pain. Although the research for establishing
4. Discussion Table 4
There are three major findings associated with this study. First, tonsillectomy causes considerable pain which typically lasts more than 7 days. Second, post-tonsillectomy pain is poorly managed. Third, poor pain management leads to increased utilization of health services. 4.1. Pain On average, post-tonsillectomy pain followed a trajectory of intense or moderately intense pain for the first 3 days. If the trend of pain reduction followed the pattern established in the first 3 days, we could expect mean pain to have been zero by Day 6. Instead pain leveled off on Days 4–6, before
Frequencies of outcomes based on daily interviews with parents Outcome Throat pain Ear pain Poor fluid intake Poor food intake Not sleeping through night Nausea Vomiting Jaw pain Minor bleeding Total complaints
Frequency 439 153 145 117 116 100 59 50 39 1218
44
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
clinically meaningful pain scores has not been done, we can agree that a pain score of 30 mm suggests problems with post-tonsillectomy pain management. 4.2. Management To achieve satisfactory pharmacological pain management, three elements must be achieved: the correct drug must be prescribed, the dose must be correct, and it must be administered in an appropriate manner. The prescription is the responsibility of the physician and the administration the responsibility of the parent. Nurses, physicians and pharmacists are involved in the education of parents and children about drug administration. Most children in this study were prescribed acetaminophen plain (the drug of choice for mild but not moderate or severe pain). We noted that many analgesic prescriptions were less than the therapeutic dose recommended by the International Association for the Study of Pain (1992). As well, analgesics were prescribed on a ‘prn’ basis rather than at regular intervals. Comments made by some parents indicated that they interpreted ‘prn’ to mean that the drug was to be administered only if pain became severe. Other parents did administer drugs regularly during waking hours but not at night time even though the child was awakened to be given antibiotics. The analysis of drug administration data is complex whether the drug and dose are controlled in an investigation or not. In this case, several analgesics at various dosages were prescribed creating many drug-dose combinations. Our analysis using percentage of daily dose administered does not take errors in dose or type of drug into consideration but only reflects willingness of parents to give the analgesic. Not having a large enough sample for many drug-dose combinations, we were unable to examine the benefits of one regimen over another. We believe that this is an important area for future investigation. Parents gave a small percentage of total analgesics that could have been administered. Parents who made visits to medical practitioners because of concern about pain, gave only half of the prescribed dose on average. Some parents commented that an analgesic was ineffective but did not ask the physician for a new prescription, nor did they provide the full prescribed dose. Another parent gave almost twice the recommended dose rather than seeking a change in prescription. Thus, parents did not administer analgesics appropriately and they were not very knowledgeable about analgesics. This indicates that physicians, nurses and pharmacists may not be fulfilling their obligation regarding education of parents and children about pain management. A gap in parent education may have led to the anomalous finding regarding long-acting infiltrations. We expected by the evening of surgery that the infiltration group would report similar pain intensity as the no-infiltration group. Instead, their pain increased as the day wore on and was greater than the no-infiltration group by the evening of sur-
gery. We suggest that the explanation for this finding lies in parents’ belief that the infiltration abolished pain or in their interpretation of the child’s early response to tonsillectomy. Parents may have interpreted an early lack of pain complaints in children who had infiltration to mean that the child had weathered the procedure well and would have a rapid recovery. These early observations may have established their pattern of postoperative care. We found consistency in amount of analgesics given day-by-day. Parents who started off administering low doses of analgesics tended to administer low doses thereafter. Thus, the initial edge provided by the infiltration was lost when the child’s behavior did not encourage parents to provide analgesics early and regularly during recovery. It is apparent therefore that in order to achieve optimal postoperative analgesic treatment, health education must be started early and it must provide the details parents require in order to treat pain knowledgeably. This is especially apparent for children who do not exhibit or complain of pain early in the course of the recovery. 4.3. Pain and utilization of health services This study offers evidence that unresolved postoperative pain is associated with increased utilization of health services and in some cases re-hospitalization. At least two cases raise suspicions that poor pain management led to re-hospitalization (prior to admission, both children reported substantially more pain than other children and were unwilling to drink and eat). For many other cases, parents visited a physician because of the severity and duration of the child’s pain. The children who obtained medical consultations reported more severe pain after Day 3 compared to those who did not obtain consultation. While it is possible that pain experienced by those who utilized health services was caused by postoperative complications, our data suggest otherwise. No complications were identified by the physicians according to parent reports. Moreover, there is evidence that pain was not well-managed. Although concerned about pain, parents who took their child to a physician provided only about half of the prescribed analgesic on Days 5, 6 and 7. Although precise costs of inadequately managed pain were not assessed in this study, we can guess that they were considerable based on the number of unscheduled medical consultations made by parents to family practitioners and to Emergency. In our region, the goal of achieving fiscal responsibility in health care has often signified the end to hospital pre-admission programs. Unfortunately, this study indicates that savings in one area (reduction in parent and child education programs) may raise costs in other areas (increase in physician consultations). Therefore, it is important that parent and child education programs be improved or maintained in light of the trend of conducting many pediatric surgeries in day surgery. Parents
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
will require more skills and knowledge about pain management if they are to manage their child’s postoperative pain successfully.
Acknowledgements This study was made possible by the support of the following: Alberta Heritage Foundation for Medical Research, Canadian Pain Society, Caritas Health Group, National Health and Research Development Program and Toronto Hospital for Sick Children-Canadian Nurses Foundation.
References Benestad, B., Vinge, O., Veierod, M.B. and Vandvik, I.H., Quantitative and qualitative assessment of pain in children with juvenile chronic arthritis based on the Norwegian version of the Pediatric Pain Questionnaire, Scand. J. Rheumatol., 25 (1996) 293–299. Beyer, J., DeGood, D., Ashley, L. and Russell, G., Patterns of postoperative analgesic use with adults and children following cardiac surgery, Pain, 17 (1983) 71–81. Carabott, J.A., Javaheri, Z., Keilty, K. and Manger, G., Oral fluid intake in
45
children following tonsillectomy and adenoidectomy, Pediatr. Nursing, 18 (1992) 124–127. Doherty, E., Yanni, G., Conroy, R.M. and Bresnihan, B., A comparison of child and parent ratings of disability and pain in juvenile chronic arthritis, J. Rheumatol., 20 (1993) 1563–1566. Finley, G.A., McGrath, P.J., Forward, S.P., McNeill, G. and Fitzgerald, P., Parents’ management of children’s pain following ‘minor’ surgery, Pain, 64 (1996) 83–87. Forward, S.P., Bowen, T.L. and McGrath, P.J., Mothers’ attitudes and behavior toward medicating children’s pain, Pain, 67 (1996) 469–474. Gabalski, E.C., Mattucci, K.F., Setzen, M. and Moleski, P., Ambulatory tonsillectomy and adenoidectomy, Laryngoscope, 106 (1996) 77–80. Gedaly-Duff, V. and Ziebarth, D., Mothers’ management of adenoid-tonsillectomy pain in 4 to 8 year olds: a preliminary study, Pain, 57 (1994) 293–299. Holt, G.R., Watkins, T.M., Yoder, M.G. and Garcia, A., The effect of tonsillectomy on impedance audiometry, Otolaryngol. Head Neck Surg., 89 (1981) 20–26. International Association for the Study of Pain, Management of Acute Pain: A Practical Guide, IASP Publications, Seattle, WA, 1992, p. 36. Lander, J., Hodgins, M. and Fowler-Kerry, S., Children’s pain predictions and memories, Behav. Res. Ther., 30 (1992) 117–124. Mather, L. and Mackie, J., The incidence of postoperative pain in children, Pain, 15 (1983) 271–282. McGrath, P.A., de Veber, L. and Hearn, M., Multidimensional pain assessment in children. In: H. Fields, R. Dubner and F. Cervero (Eds.), Advances in Pain Research and Therapy, Raven Press, New York, 1985, pp. 387–393.
Pain progression, intensity and outcomes following tonsillectomy Fay F. Warnock*, Janice Lander 3-132 Clinical Sciences Building, University of Alberta, Edmonton, Alberta, T6G 2G3, Canada Received 4 August 1997; received in revised form 8 October 1997; accepted 7 November 1997
Abstract The objective of this study was to assess outcomes of pediatric day surgery tonsillectomy. A total of 129 children, aged 5–16 years, and their parents were recruited from three urban hospitals which provided pediatric day surgery. Children reported pain on a visual analogue scale (VAS) in day surgery and then daily at home for 7 days. Parents reported outcomes of surgery, including fluid intake, nausea, vomiting and sleep disturbances. They also recorded analgesic administration. Three main results related to extent and duration of pain, quality of management of pain, and effect of pain on utilization of health services. Tonsillectomy caused considerable pain which lasted more than 7 days. Pain followed a trajectory of intense or moderately intense pain for the first 3 days followed by a gradual decline over the next 4 days. In general, post-tonsillectomy pain was poorly managed by health professionals and parents. An unexpected observation was that children who had a bupivacaine infiltration of the tonsil fossa during surgery had significantly more pain in the evening of surgery than children who did not have an infiltration. The increase in postoperative pain experienced by those who had the infiltration was attributed to quality of pain management. Children with persistent pain (those who did not follow the typical trajectory) were likely to be taken to a medical practitioner. One-third of the sample made unscheduled visits to practitioners with most occurring from Day 4 to Day 7 of the follow-up. 1998 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Tonsillectomy; Postoperative pain; Day surgery; School-aged child; Adolescents
1. Introduction Researchers have been aware for some time that children’s postoperative pain is managed poorly (Beyer et al., 1983; Mather and Mackie, 1983). Perhaps one explanation relates to a lack of information about prevalence and patterns of acute postoperative pain caused by surgical procedures, including tonsillectomy. Tonsillectomy is one of the most common of pediatric surgeries. In several studies that have been conducted, parents or staff have rated children’s postoperative pain (Gedaly-Duff and Ziebarth, 1994; Finley et al., 1996; Gabalski et al., 1996). According to parents, children’s post-tonsillectomy pain is moderately high on the first 2 days following surgery (Finley et al., 1996). Parent ratings, however, are not always in agreement with the child’s rating (see for example, Doherty et al., 1993; Benestad et al.,
1996). Despite their assessment of postoperative pain, parents under-administered analgesics (Gedaly-Duff and Ziebarth, 1994; Finley et al., 1996). Parents’ fears about tolerance, addiction and drug side-effects and their lack of knowledge about pain management are barriers to effective administration of analgesics (Gedaly-Duff and Ziebarth, 1994; Finley et al., 1996; Forward et al., 1996). Since many tonsillectomies are performed in day surgery, postoperative care including pain-management is being provided by parents in the home. To determine intensity and progression of pain as well as pain outcomes, we undertook a 7-day follow-up of children who had tonsillectomy in day surgery.
2. Methods 2.1. Subjects and setting
* Corresponding author. Tel.: +1 403 4312051; fax: +1 403 4922551; e-mail: [email protected]
This study was conducted in all hospitals in a mid-west
0304-3959/98/$19.00 1998 International Association for the Study of Pain. Published by Elsevier Science B.V. PII S0304-3959 (97 )0 0202-9
38
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Canadian city which provide pediatric day surgery for tonsillectomy for the area (n = 3 hospitals). The region is multi-cultural. Most of the children having surgery live in the city or within 40 km of it. Others come from outside of the city, some traveling as much as 800 km for surgery. All children, 5–16 years, having tonsillectomy day surgery were recruited during the 13-week sampling phase of the study. An additional inclusion criterion was that children be able to use a visual analogue scale. Surgical procedures performed on the study sample consisted of tonsillectomy alone or tonsillectomy with adenoidectomy and/or myringotomy. All procedures were performed by otolaryngologists or general surgeons and under general anesthesia. In addition, tonsil fossa of some children were infiltrated with bupivacaine HCl 0.25% or 0.50% with or without epinephrine during surgery. 2.2. Instruments A 100 mm visual analogue scale (VAS) was used for assessing pain intensity. The VAS is a valid, reliable and easily understood tool which has been used successfully to obtain self-reported pain from children ≥5 years of age (McGrath et al., 1985). The VAS was arranged as a 100 mm vertical line with anchors at either end (No Pain and Worst Pain Possible). Children’s comprehension and proficiency at using the VAS were assessed with an approach utilized in other research (Lander et al., 1992). Instructions about use of the VAS were provided to children who were then required to correctly rank sketched faces that illustrated pain. No child was excluded from this study because of inability to use the VAS. 2.3. Procedure The study procedure was divided into three phases. These included: pre-operative period, postoperative period while the child was in day surgery, and postoperative period at home. 2.3.1. Pre-operative phase Shortly after their arrival at the day surgery, the study was described to parents of children having tonsillectomy. If the parent agreed to participate, the child was recruited and instructed about the use of the VAS. The child’s ability to use the tool was then assessed. Parents provided information about the child’s general health status and history of previous hospitalizations. They were instructed about use of a medication diary. An envelope which contained VAS scales and a medication diary was provided for use at home. 2.3.2. Postoperative day surgery phase After tonsillectomy, children recovered in the day surgery for a period of time. Two to three hours after surgery,
children were asked to rate their current pain using the VAS. At the end of this phase, the research assistant recorded information from the hospital chart about the surgical procedure and prescribed analgesics. 2.3.3. Postoperative home phase Beginning with the day of surgery (Day 1) and every evening for 6 additional days (Days 2–7), parents were contacted by telephone in the early evening. On Day 1 parents reported for the period of time from when they returned home until the interview. For Days 2–7, they reported for the 24 h period of time since the previous interview. During each phone call, which took approximately 5– 10 min, parents were asked if the child had or had not experienced any of the following: pain; nausea and vomiting; postoperative bleeding; and uninterrupted night time sleep. They were also asked for the amount of fluid taken by the child. At the end of the interview, they were given the opportunity to describe any other concerns related to the child’s progress and care. After the parent was interviewed each evening, the research assistant asked to have the child brought to the telephone. The child was re-instructed about use of the VAS and asked to report current pain by making a pencil mark across the vertical line. A new VAS was used for each day’s pain report and the child was instructed to place the completed and dated VAS in the storage packet that had been provided. At the end of the 7-day follow-up period, parents were asked to return the VAS scales and medication log in the stamped and addressed envelope that had been provided. 2.3.4. Data preparation All data were coded and examined for completeness and accuracy. Distributions were inspected and descriptive statistics obtained for all variables. VAS pain ratings were computed by measuring the distance along the line to the point where the subject’s mark intersected the line. Parent reports of outcomes such as pain, nausea, vomiting, sleeping and bleeding, as reported in the telephone interview, were coded as dichotomous variables. Fluid intake was used as an indicator of risk of dehydration. Since others (Carabott et al., 1992) have set the minimum acceptable level of postoperative intake at 500 ml over 8 h, we coded an intake of ,30 ml/h as poor. Frequencies of all parent-reported outcomes were computed. A variable referred to as ‘percentage of daily dose administered’ was computed as follows: the total milligrams of analgesic administered was divided by the maximum amount of analgesic that could be administered, and the result was then multiplied by 100. This variable was computed separately for Days 2–7 for 108 cases (no dose had been specified for the other four cases). Day 1 was not included since the period of time that parents could have administered analgesics was variable depending on the time of surgery.
39
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
3. Results 3.1. Subject recruitment A total of 147 children, aged 5–16 years, had tonsillectomy during the recruitment interval. All were approached for study participation with the exception of five children who were taken to the operating room earlier than anticipated, before the arrival of the research assistant. Of those 142 approached, 10 did not participate for the following reasons: four children declined although their parents had consented and six parents declined because they lacked time or were reluctant to have their child involved in research. Therefore, 132 children were entered into the study. Two children were lost to follow-up when their families could not be reached by telephone (thus, n = 130). The number of cases available for analysis depended on the type of data being analyzed. Six children were admitted to hospital at some point during follow-up thus reducing the sample size for child-reported pain data and parent telephone interview data. One family did not complete any interviews as the child was admitted directly from day surgery. Several telephone interviews were conducted with four parents before or after the child was admitted. For one parent and child, all interviews and pain assessments were conducted as admission occurred on the last day of follow-up (this case was included in data analysis). The sample was reduced to n = 125 for pain data but n = 129 for telephone interview data. The VAS scales and drug administration logs were not returned or went missing in the mail for 13 subjects who completed all telephone interviews (n = 112). Therefore, analysis of the interview data was conducted with data from 129 cases whereas analysis of pain data was conducted with 112 cases. 3.2. Subject characteristics The children of parents who were interviewed (n = 129) had a mean age of 9.3 years (SD = 3.5). There were 64 girls and 65 boys. Approximately one third of these children had not been hospitalized previously and most were having their first surgery. Further details about characteristics of the sample are reported in Table 1. The rate of surgical procedures performed were: tonsillectomy with adenoidectomy (n = 81, 63%), tonsillectomy alone (n = 39, 30%), tonsillectomy with myringotomy and adenoidectomy (n = 9, 7%). No further analyses of surgical procedures were done because of insufficient numbers in two categories. 3.3. Trajectory of postoperative pain During the postoperative day surgery phase, the mean VAS score was 66.6 (SD = 32.7, n = 120). Nine young children (mean age 5.7 years) refused to rate pain while recovering in day surgery.
A repeated measures analysis of variance was used to assess changes in postoperative pain as reported by the children on the telephone for Day 1 to Day 7 (F = 70.7, P , 0.0001, n = 112). There was a significant drop in pain from Day 1 to Day 7. Mean daily ratings of postoperative pain intensity ranged from 64.8 (Day 1: SD = 29.2) to 26.8 (Day 7: SD = 25.0). Fig. 1 illustrates daily mean VAS scores. Frequencies of children reporting VAS pain ratings of ≥30 mm were determined for each of eight pain measures obtained. The frequencies were: Day Surgery, 97 (80.8%); Day 1, 99 (88.4%); Day 2, 84 (75.0%); Day 3, 69 (61.6%); Day 4, 51 (45.5%); Day 5, 51 (45.5%); Day 6, 47 (42.0%); and Day 7, 34 (30.4%). 3.4. Prescribed and administered analgesics Analgesics prescribed for postoperative pain were obtained from subject charts (n = 129). The drugs prescribed were: acetaminophen plain, n = 77 (60%); acetaminophen with codeine, n = 34 (26%); and codeine, n = 18 (14%). Percentage of daily doses administered were computed for analgesics as described previously. Correlations between daily child-reported pain and percentages of daily dose administered were near zero and non-significant except for Day 7 (r = 0.22, P = 0.02). By contrast, the correlations among percentage of daily dose administered were moderate to strong positive and significant. For example, the percentages administered in the first 2 days were correlated with the percentage administered on subsequent days (Day 2 with Day 3–7 correlations ranged from 0.37 to 0.66; and Day 3 with Day 4–7 correlations ranged from 0.59 to 0.79). Table 1 Subject characteristics (n = 129) Participants completing telephone interviews (%) Gender Boys Girls Past hospitalizations None Once More than once Previous surgeries None Once More than once General health status Excellent Very good Good Poor Residence Urban Rural
65 (50.4) 64 (49.6) 52 (40.3) 42 (32.6) 35 (27.1) 86 (66.7) 32 (24.8) 11 (8.5) 25 61 28 15
(19.4) (47.3) (21.7) (11.6)
101 (78.3) 28 (21.7)
40
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Receiving a low percentage of daily dose on Day 2 or 3 was therefore associated with receiving a low dose on subsequent days. Few children received the full dose each day, including on the days immediately following surgery. In fact on the day following surgery, half of the children received less than 50% of the analgesics that they could have received and only nine (8.3%) received the full dose. Several children received more than prescribed including one 5 year old who was given almost twice the prescribed and recommended dose of acetaminophen on 5 days of the follow-up. Night-time analgesic administration patterns were also examined for Day 1 to Day 7. On 4 of 7 nights, including the day of surgery, only 25% of the sample received analgesics at night-time. On the second day, about half of the sample were given analgesics during the night and about one-third during the third and fifth nights. Bupivacaine HCl was infiltrated at the tonsil fossa for 35 children (27%) who also had a general anesthetic. Their pain during follow-up was compared with those who did not receive infiltration. We were interested in examining overall differences in the two groups occurring over follow-up days (group by time interaction). Age, however, was correlated with child-reported pain (r = 0.20) and those who received tonsil fossa local infiltration were significantly older (mean 11.2 years, SD = 3.6) than those who did not (mean 8.8 years, SD = 3.3) (F = 11.0, P = 0.001). Thus, age was related to both grouping and outcome variables posing difficulties for data interpretation. We therefore matched each child who received a local infiltration with a child of the same age who did not receive a local infiltration. This strategy created a reduced sample of two
age-matched groups of n = 27 each, thereby controlling age. Equivalence of the matched groups was confirmed for gender, age and child’s weight (with analysis of variance or chi-square). Repeated measures analysis of variance indicated significant differences in pain for the two infiltration groups depending on follow-up time (F = 2.2, P = 0.04). Fig. 2 illustrates child-reported pain for the two groups over the follow-up period. Those who did not have the infiltration had a progressive decrease in pain over the follow-up period whereas the pain of those having the bupivacaine infiltration rose during the evening of surgery before progressively declining over the next 5 days. The infiltration group had significantly less pain in day surgery and more pain during the evening compared to the no-infiltration group (Day Surgery, infiltration group 64.8, no-infiltration group 80.3; Evening of Surgery, infiltration group 75.9, no-infiltration group 67.2). Percentage of analgesics were compared from Day 2 to Day 7 for the two age-matched infiltration groups (with repeated measures analysis of variance). There was a significant interaction between infiltration group and time (F = 6.0, P , 0.001). The no-infiltration group received significantly more analgesics on Days 2 to 4 compared to the infiltration group (Fig. 3). For example, on Day 2 the noinfiltration group received 63.5% of prescribed analgesics compared to 33.8% for the infiltration group. The age-matched groups were compared and significant differences in type of prescribed drug found (chisquare = 8.0, P = 0.02). The no-infiltration group was prescribed more acetaminophen plain and the infiltration group more codeine and more acetaminophen with codeine (Table 2).
Fig. 1. Mean pain ratings over 7 days (n = 112) 1 (DS) refers to VAS in day surgery on Day 1, 1 (Home) refers to VAS at home on Day 1.
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
41
Fig. 2. Mean pain for Infiltration and No-Infiltration groups. Groups matched by age (n = 27). 1 (DS) refers to VAS in day surgery on Day 1, 1 (Home) refers to VAS at home on Day 1.
3.5. Common postoperative occurrences As mentioned, six parents reported postoperative complications requiring admission to hospital. Two children were admitted directly from day surgery due to postoperative hemorrhage and three other children were hospitalized on the second postoperative day, two for ear and throat infection and one for throat pain and infection. Another was hospitalized on the seventh postoperative day for dehydra-
tion and hypoglycemia secondary to refusal to swallow fluids and food. Parents also reported occurrences of any medical consultations that were not part of pre-arranged postoperative care. Forty-nine (38.3%) parents reported having to make unscheduled medical visits or phone calls related to the child’s recovery from surgery. Forty required a single unscheduled consultation during the 7-day follow-up whereas the other nine required multiple consultations
Fig. 3. Mean percent of analgesics for Infiltration and No-Infiltration groups. Groups matched by age (n = 25)
42
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Table 2
Table 3
Frequencies of analgesics prescribed for age-matched infiltration and noinfiltration groups
Type and frequency of unscheduled medical consultations.
Acetaminophen plain (%)
Acetaminophen with codeine (%)
No-infiltration Infiltration
16 (59.3) 6 (22.2)
8 (29.6) 13 (48.1)
Total
22
21
Codeine (%) 3 (11.1) 8 (29.6)
Frequency of consultation Family physician Surgeon Walk-in clinic Phoned day surgery Emergency
37 7 6 6 5
Total
61
11
(two consultations, n = 6; three consultations, n = 3). Therefore, a total of 61 consultations were obtained. The majority (74%) of consultations were obtained on Days 4–7 (with 66% taking place on Days 5–7). Table 3 describes type of consultations made. Those who did and did not obtain unscheduled medical consultations were compared. The two groups were not significantly different for the following: (i) child’s health status, gender, history of previous hospitalizations or surgeries, whether or not parents had previous experience with a child in day surgery, whether they resided in an urban or rural area, and type of prescribed analgesic (using chi-square analysis); and (ii) parent’s expressed worry about caring for the child at home on a scale of 1 to 10 (using analysis of variance). A significant repeated measures analysis of covariance was obtained when consultation and no-consultation groups were compared for child-reported pain on Days 1–7 (child’s age was the covariate). The progression of pain for the no-
consultation group followed the trajectory plotted for the total group (as noted in Fig. 1). For the consultation group, pain declined initially, leveled off on Day 4 and then increased beginning on Day 5 (Fig. 4; F = 4.0, P = 0.001). Children of parents who required unscheduled consultations had significantly more pain, twice as much on average, on Days 5, 6 and 7. The two groups of parents who did or did not seek unscheduled consultations were compared also for percentage of drugs administered on Days 2–7 with repeated measures analysis of covariance (child’s age as covariate). There were overall significant differences in percent of analgesics administered for the two groups of parents (F = 5.5, P = 0.02). Mean overall percent administered for the no-consultation was 27.5% and 51.2% for the consultation group. As well, there was a significant interaction between consultation group and follow-up day (Fig. 5; F = 7.2, P , 0.001). On Days 5, 6 and 7, parents who
Fig. 4. Mean daily pain for groups obtaining or not obtaining medical consultation.
43
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
Fig. 5. Mean percentage of drugs administered for groups obtaining or not obtaining medical consultation.
sought a medical consultation gave a significantly greater percentage of analgesics (52.8%, 53.5% and 54.0%, respectively) compared to parents who did not seek consultation (18.2%, 14.8% and 11.6%, respectively). Parents reported various outcomes during the telephone interview. Since parents could report multiple outcomes on any day and repeat the outcome from day to day, total frequencies for outcomes are reported in Table 4. Of note are the findings of inadequate fluid intake (which were reported during the first few days), and the occurrence of ear and jaw pain (which were reported increasingly in the latter days of the follow-up period).
beginning a decline again. The change in the downward trend of pain at Day 4 coincided with reports of onset of ear and jaw pain. Some clinicians have suggested that surgical trauma and postoperative inflammation affects coordination of nasopharyngeal and eustachian tube muscles (Holt et al., 1981). This may be one explanation for occurrence of ear and jaw pain. An interruption in the downward trajectory of throat pain perhaps stems from the return to a normal diet (with concomitant increase in swallowing) as the child convalesces. On Day 7, when pain was at its lowest, almost one-third of children rated pain as 30 mm or more on the VAS. This was the criterion used by Finley et al. (1996) for clinically significant pain. Although the research for establishing
4. Discussion Table 4
There are three major findings associated with this study. First, tonsillectomy causes considerable pain which typically lasts more than 7 days. Second, post-tonsillectomy pain is poorly managed. Third, poor pain management leads to increased utilization of health services. 4.1. Pain On average, post-tonsillectomy pain followed a trajectory of intense or moderately intense pain for the first 3 days. If the trend of pain reduction followed the pattern established in the first 3 days, we could expect mean pain to have been zero by Day 6. Instead pain leveled off on Days 4–6, before
Frequencies of outcomes based on daily interviews with parents Outcome Throat pain Ear pain Poor fluid intake Poor food intake Not sleeping through night Nausea Vomiting Jaw pain Minor bleeding Total complaints
Frequency 439 153 145 117 116 100 59 50 39 1218
44
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
clinically meaningful pain scores has not been done, we can agree that a pain score of 30 mm suggests problems with post-tonsillectomy pain management. 4.2. Management To achieve satisfactory pharmacological pain management, three elements must be achieved: the correct drug must be prescribed, the dose must be correct, and it must be administered in an appropriate manner. The prescription is the responsibility of the physician and the administration the responsibility of the parent. Nurses, physicians and pharmacists are involved in the education of parents and children about drug administration. Most children in this study were prescribed acetaminophen plain (the drug of choice for mild but not moderate or severe pain). We noted that many analgesic prescriptions were less than the therapeutic dose recommended by the International Association for the Study of Pain (1992). As well, analgesics were prescribed on a ‘prn’ basis rather than at regular intervals. Comments made by some parents indicated that they interpreted ‘prn’ to mean that the drug was to be administered only if pain became severe. Other parents did administer drugs regularly during waking hours but not at night time even though the child was awakened to be given antibiotics. The analysis of drug administration data is complex whether the drug and dose are controlled in an investigation or not. In this case, several analgesics at various dosages were prescribed creating many drug-dose combinations. Our analysis using percentage of daily dose administered does not take errors in dose or type of drug into consideration but only reflects willingness of parents to give the analgesic. Not having a large enough sample for many drug-dose combinations, we were unable to examine the benefits of one regimen over another. We believe that this is an important area for future investigation. Parents gave a small percentage of total analgesics that could have been administered. Parents who made visits to medical practitioners because of concern about pain, gave only half of the prescribed dose on average. Some parents commented that an analgesic was ineffective but did not ask the physician for a new prescription, nor did they provide the full prescribed dose. Another parent gave almost twice the recommended dose rather than seeking a change in prescription. Thus, parents did not administer analgesics appropriately and they were not very knowledgeable about analgesics. This indicates that physicians, nurses and pharmacists may not be fulfilling their obligation regarding education of parents and children about pain management. A gap in parent education may have led to the anomalous finding regarding long-acting infiltrations. We expected by the evening of surgery that the infiltration group would report similar pain intensity as the no-infiltration group. Instead, their pain increased as the day wore on and was greater than the no-infiltration group by the evening of sur-
gery. We suggest that the explanation for this finding lies in parents’ belief that the infiltration abolished pain or in their interpretation of the child’s early response to tonsillectomy. Parents may have interpreted an early lack of pain complaints in children who had infiltration to mean that the child had weathered the procedure well and would have a rapid recovery. These early observations may have established their pattern of postoperative care. We found consistency in amount of analgesics given day-by-day. Parents who started off administering low doses of analgesics tended to administer low doses thereafter. Thus, the initial edge provided by the infiltration was lost when the child’s behavior did not encourage parents to provide analgesics early and regularly during recovery. It is apparent therefore that in order to achieve optimal postoperative analgesic treatment, health education must be started early and it must provide the details parents require in order to treat pain knowledgeably. This is especially apparent for children who do not exhibit or complain of pain early in the course of the recovery. 4.3. Pain and utilization of health services This study offers evidence that unresolved postoperative pain is associated with increased utilization of health services and in some cases re-hospitalization. At least two cases raise suspicions that poor pain management led to re-hospitalization (prior to admission, both children reported substantially more pain than other children and were unwilling to drink and eat). For many other cases, parents visited a physician because of the severity and duration of the child’s pain. The children who obtained medical consultations reported more severe pain after Day 3 compared to those who did not obtain consultation. While it is possible that pain experienced by those who utilized health services was caused by postoperative complications, our data suggest otherwise. No complications were identified by the physicians according to parent reports. Moreover, there is evidence that pain was not well-managed. Although concerned about pain, parents who took their child to a physician provided only about half of the prescribed analgesic on Days 5, 6 and 7. Although precise costs of inadequately managed pain were not assessed in this study, we can guess that they were considerable based on the number of unscheduled medical consultations made by parents to family practitioners and to Emergency. In our region, the goal of achieving fiscal responsibility in health care has often signified the end to hospital pre-admission programs. Unfortunately, this study indicates that savings in one area (reduction in parent and child education programs) may raise costs in other areas (increase in physician consultations). Therefore, it is important that parent and child education programs be improved or maintained in light of the trend of conducting many pediatric surgeries in day surgery. Parents
F.F. Warnock, J. Lander / Pain 75 (1998) 37–45
will require more skills and knowledge about pain management if they are to manage their child’s postoperative pain successfully.
Acknowledgements This study was made possible by the support of the following: Alberta Heritage Foundation for Medical Research, Canadian Pain Society, Caritas Health Group, National Health and Research Development Program and Toronto Hospital for Sick Children-Canadian Nurses Foundation.
References Benestad, B., Vinge, O., Veierod, M.B. and Vandvik, I.H., Quantitative and qualitative assessment of pain in children with juvenile chronic arthritis based on the Norwegian version of the Pediatric Pain Questionnaire, Scand. J. Rheumatol., 25 (1996) 293–299. Beyer, J., DeGood, D., Ashley, L. and Russell, G., Patterns of postoperative analgesic use with adults and children following cardiac surgery, Pain, 17 (1983) 71–81. Carabott, J.A., Javaheri, Z., Keilty, K. and Manger, G., Oral fluid intake in
45
children following tonsillectomy and adenoidectomy, Pediatr. Nursing, 18 (1992) 124–127. Doherty, E., Yanni, G., Conroy, R.M. and Bresnihan, B., A comparison of child and parent ratings of disability and pain in juvenile chronic arthritis, J. Rheumatol., 20 (1993) 1563–1566. Finley, G.A., McGrath, P.J., Forward, S.P., McNeill, G. and Fitzgerald, P., Parents’ management of children’s pain following ‘minor’ surgery, Pain, 64 (1996) 83–87. Forward, S.P., Bowen, T.L. and McGrath, P.J., Mothers’ attitudes and behavior toward medicating children’s pain, Pain, 67 (1996) 469–474. Gabalski, E.C., Mattucci, K.F., Setzen, M. and Moleski, P., Ambulatory tonsillectomy and adenoidectomy, Laryngoscope, 106 (1996) 77–80. Gedaly-Duff, V. and Ziebarth, D., Mothers’ management of adenoid-tonsillectomy pain in 4 to 8 year olds: a preliminary study, Pain, 57 (1994) 293–299. Holt, G.R., Watkins, T.M., Yoder, M.G. and Garcia, A., The effect of tonsillectomy on impedance audiometry, Otolaryngol. Head Neck Surg., 89 (1981) 20–26. International Association for the Study of Pain, Management of Acute Pain: A Practical Guide, IASP Publications, Seattle, WA, 1992, p. 36. Lander, J., Hodgins, M. and Fowler-Kerry, S., Children’s pain predictions and memories, Behav. Res. Ther., 30 (1992) 117–124. Mather, L. and Mackie, J., The incidence of postoperative pain in children, Pain, 15 (1983) 271–282. McGrath, P.A., de Veber, L. and Hearn, M., Multidimensional pain assessment in children. In: H. Fields, R. Dubner and F. Cervero (Eds.), Advances in Pain Research and Therapy, Raven Press, New York, 1985, pp. 387–393.