Table 26, KQ3 monitoring strategies evidence - ADHD Diagnosis and Treatment in Children and Adolescents

Table 26KQ3 monitoring strategies evidence

Study: Author, Year; Location	Intervention, Analysis, Follow Up	Results
Cedergren, 2021¹⁷³ Sweden	Open-label monitoring consisting of 5 follow-up visits in 12 months using a continuous performance test (QbTest) and investigator rating on the ADHD-RS. Qualitative comparison of change in ADHD-RS and QbTest scores over 12 months Naturalistic follow up, with medication administered according to clinician judgement of need.	Bonferroni-adjusted pairwise comparisons showed statistically significant reductions in QbTest and ADHD-RS scores over the 12-month study. Both measures appear to capture symptom change over time, but weak correlations between the measures suggest that their role in medical follow-up might be complementary rather than interchangeable.
Cohen, 1989²⁰³ United States	Randomized, double-blind, placebo-controlled crossover study of the use of monitoring ADHD symptoms – before and during treatment with methylphenidate – using the ADD-H Comprehensive Teacher Rating Scale, Conners parent rating scale, and the Gordon Diagnostic System (a computerized continuous performance task assessing vigilance and impulse control). Group differences in change in symptom scores over time. Naturalistic follow up, before and during treatment with fixed-dose (5mg for children weighing less than 30kg, 10mg for children weighing 30kg or more), short-acting methylphenidate administered twice daily for 1 month, with measures collected at baseline, 1 month (the time of crossover), and 2 months (endpoint).	Both rating scales demonstrated significant change in symptoms (inattention and hyperactivity on the ADD-H scale; hyperactivity on the Conners scale) during treatment with methylphenidate compared with placebo, whereas the Gordon task did not demonstrate change. Rating scales, but not this continuous performance task, appear helpful in monitoring the short-term effects of stimulant treatment.
Epstein, 2007²⁵⁶ United States	12 pediatric practices were randomly assigned to receive access to collaborative consultative services or a control group. In the collaborative consultation services, pediatricians were encouraged and assisted to use rating scales for symptom monitoring and titration trials to determine optimal medication dosages. Physicians were taught to prescribe 4 different doses of methylphenidate during a titration trial (placebo, 18 mg, 36 mg, 54 mg); the order of week-long dosing was blinded but standardized across patients (week 1, 18 mg; week 2, placebo; week 3, 36 mg; week 4, 54 mg) to determine optimal dosing for each patient. Parents and teachers completed weekly behavioral ratings (Conners Global Index) & side effect rating scales. Data were returned to Duke Univ psychiatrist to determine the best starting medication dose; a report describing the titration results was faxed back to pediatricians. Patients in control group practices received treatment as usual, without access to consultative services. Assessed Conners Global Index & side effect rating scales. Monthly follow up with Conners and side effect rating scales for 12 months, sent to Duke U psychiatrists for interpretatin, with recommendations returned to the pediatrician	Use of symptom ratings did not differ significantly by group, nor did the change in symptoms over time. Pediatrician compliance with the collaborative consultation service was poor (pediatricians for 29 of 59 patients in the consultation group received a titration trial and 13/59 participated in monthly medication monitoring). Preliminary secondary analyses indicated that those children whose pediatricians complied with titration had significantly better outcomes compared with those who did not and TAU controls (group × time P<.01) Children in the collaborative consultation service–complier group had a 27% reduction in symptom scores compared with 18% reduction in the TAU controls and 13% reduction in consultation non-compliers.
Epstein, 2016²⁵⁵ United States	Cluster randomized controlled trial of either a technology-assisted quality improvement (QI) intervention or TAU control. QI intervention consisted of 4 training sessions, office flow modification, guided QI, and an ADHD Internet portal to assist with treatment monitoring versus TAU control practices Assessed intervention effects on parent- and teacher-rated ADHD severity using on the Vanderbilt ADHD total symptom score. 12 months follow up	Intent-to-treat analyses examining outcomes (parent ratings of ADHD severity) in all 577 children assessed for ADHD were not significant (b=−1.97, P=0.08), but among the 373 children prescribed ADHD medication, a significant intervention effect on reducing parent-rated symptom severity (b=−2.42, P=0.04) but not teacher-rated symptoms was observed. Prescriber compliance with treatment guidelines was poor, as only 373 of the 577 patients received medication at any time in the 1-year follow-up, and many who did receive it were prescribed sub-optimal doses. Compared with the usual care group, providers in the intervention group had 25% more patient contacts (d=.38, p=.0008) and collected 4.6 (d=.57, p<.0001) and 9.9 (d=.54, p<.0001) times more parent and teacher ratings, respectively. However, providers in the intervention group collected parent ratings in only half and teacher ratings in a quarter of their patients during the initial year of medication treatment.
Fiks, 2017²⁶⁸ United States	Cluster-randomized open label trial at the practice level (9 intervention, 10 control sites) for 3-component quality-improvement program that employs distance learning: (1) 3 15-minute Web-based presentations on evidence-based practices for managing ADHD in primary care; (2) optional collaborative consultation with ADHD experts via a health system online networking site or private email/telephone conversation; (3) and performance feedback reports or calls every 2 months informing them of their rates of sending and receiving ADHD rating scales from parents and teachers and allowed them to compare their results to results of the entire group; feedback reports were discussed during four, 1-hour conference calls). Participation qualified for Maintenance of Certification credit from the American Board of Pediatrics. Collection of rating scales was facilitated via an electronic application linked to the electronic health record versus waitlist control Number of parent and teacher rating scales sent out and received back assessed	Differences between intervention arms were not statistically significant, though clinicians in both study arms were significantly more likely to administer and receive parent and teacher rating scales compared to an 8-month baseline period. Intervention clinicians who participated in at least one performance feedback call were more likely to send out parent rating scales than intervention clinicians who did not participate (relative difference of 14.2 percentage points, 95% CI: 0.6, 27.7. For all study outcomes, practices with the highest rates of clinician participation in the study (≥ 80%), were not superior to practices with lower rates of involvement (< 80%). Participation was low (105 of 166 invited); 42 of 53 in the intervention group completed all 3 education presentations; 30 (57%) participated in at least one feedback call, and 19 (36%) participated in all 3 components of the intervention.
Florida International University, 2010²⁷⁴ United States	Randomized to receive either osmotic release oral system-methylphenidate alone (78%) or behavioral therapy alone (22%). After 6 months, children with a decline in body mass index >0.5 z-units were randomized to 1 of 3 weight recovery treatments: (1) monthly height/weight monitoring plus daily medication; (2) drug holidays on non-school days (with monthly monitoring); or (3) daily caloric supplements (with daily medication and monthly monitoring). Standardized body weight and height assessed 18 follow-up visits over 30 months	All groups significantly increased their weight gain. Drug holidays + monitoring, caloric supplementation + monitoring, and monitoring alone all led to increased weight velocity in children taking CNS stimulants, but with no differences between groups, and no intervention led to increased height velocity. When analyzed by what parents did (versus what they were assigned to), caloric supplementation (p<0.01) and drug holidays (p<0.05) increased weight velocity more than monitoring of height and weight. Over the entire study, participants declined in standardized weight (−0.44 z-units) and height (−0.20 z-units).
Oppenheimer, 2019⁴⁶⁶ United States	Naturalistic study of a Web-based platform enabling clinicians to administer online monthly clinical questionnaires to parents and teachers for monitoring of patients remotely between visits. Trigger algorithm alerts clinicians to clinically actionable events that are documented in the medical record versus non-alert group Patients were the unit of analysis. Parent and teacher reports of current medication, medication side effects inventory, Vanderbilt ADHD Parent Rating Scale, Clinical Global Impression-Severity (CGI-S) scale, and Clinical Global Impression-Improvement (CGI-I) scale 15 months follow up	Trigger algorithms produced alerts requiring immediate review in 8% of the parent reports. Clinicians perceived 74% of alerts to be significant enough to prompt urgent follow-up with parents, suggesting a low rate of false positive alerts. Patients who generated alerts compared to those who did not had more severe ADHD symptoms (beta = 5.8, 95% CI: 3.5–8.1 [p < 0.001] in the 90 days prior to an alert, further supporting validity of the alerts.
Smith, 2000⁵⁴⁵ United States	A assessed the reliability, validity, and unique contributions of self-reports by adolescents receiving treatment for ADHD in a summer treatment program that included self-monitoring as a treatment component. Self-reported IOWA Conners Inattention/Overactivity and Oppositional/Defiant subscales, ratings of interactions with peers and staff. Assessed changes in reliability during a placebo-controlled, cross-over study of 30 mg of methylphenidate. Observed frequencies of negative behavior, rating from parents and teachers	Average reliability for the adolescent self-report across all measures was .78 (range .74-.83), similar to the reliability of .82 for counselors (range .78-.85), and significantly better than the teacher reliability of .60 (range .51-.68). Teacher and counselor ratings on the Conners changed significantly during stimulant treatment whereas adolescent self-ratings did not. The findings suggest that adolescents can provide reliable information on their symptoms, but not beyond what parents can provide. Adolescents may also be poor sources of information about the change in ADHD symptoms, but a good source of intormation about improved interactions with others in response to treatment.
Weisman, 2018⁶⁰⁹ Israel	Mobile app allows patients or their parents to report their clinical status following initiation of prescription or after changing medication dosage; purpose of the app is to facilitate communication with MD; app includes questions on severity of ADHD symptoms and potential side effects and can also function as a medication reminder Treatment as usual, without app	CGI-Severity no significant difference No significant difference on ADHD-RS, possibly due to inadequate power, Significant difference (p= 0 .008) favoring intervention group on the Clinician Rating Scale (CRS). Intervention group had significantly better adherence, as measured by pill count (p < .015).
Yang, 2012⁶²⁹ Korea	Naturalistic study of symptom monitoring and medication adherence assessed using the Medication Event Monitoring System, a bottle cap with a microprocessor that records all instances and times that the bottle is opened Patient self-report, clinician rating, pill count assessed; measure of adherence 8 weeks follow up	The rate of non-adherence was 46.2%, higher than patient self-report of 17.9%, clinician rating of 31.7%, and pill count of 12.8%. Pill count and monitoring system concordance was 0.249 (95% CI: 0.102-0.386). Self-report concordance was 0.237 (95% CI: −0.024-0.468). Non-adherent patients had more severe symptoms at baseline and inferior improvement compared with adherent patients.

Study:

Author, Year;

Location

Intervention,

Analysis,

Follow Up

Results

Cedergren, 2021¹⁷³

Sweden

Open-label monitoring consisting of 5 follow-up visits in 12 months using a continuous performance test (QbTest) and investigator rating on the ADHD-RS.

Qualitative comparison of change in ADHD-RS and QbTest scores over 12 months

Naturalistic follow up, with medication administered according to clinician judgement of need.

Bonferroni-adjusted pairwise comparisons showed statistically significant reductions in QbTest and ADHD-RS scores over the 12-month study.

Both measures appear to capture symptom change over time, but weak correlations between the measures suggest that their role in medical follow-up might be complementary rather than interchangeable.

Cohen, 1989²⁰³

United States

Randomized, double-blind, placebo-controlled crossover study of the use of monitoring ADHD symptoms – before and during treatment with methylphenidate – using the ADD-H Comprehensive Teacher Rating Scale, Conners parent rating scale, and the Gordon Diagnostic System (a computerized continuous performance task assessing vigilance and impulse control).

Group differences in change in symptom scores over time.

Naturalistic follow up, before and during treatment with fixed-dose (5mg for children weighing less than 30kg, 10mg for children weighing 30kg or more), short-acting methylphenidate administered twice daily for 1 month, with measures collected at baseline, 1 month (the time of crossover), and 2 months (endpoint).

Both rating scales demonstrated significant change in symptoms (inattention and hyperactivity on the ADD-H scale; hyperactivity on the Conners scale) during treatment with methylphenidate compared with placebo, whereas the Gordon task did not demonstrate change.

Rating scales, but not this continuous performance task, appear helpful in monitoring the short-term effects of stimulant treatment.

Epstein, 2007²⁵⁶

United States

12 pediatric practices were randomly assigned to receive access to collaborative consultative services or a control group. In the collaborative consultation services, pediatricians were encouraged and assisted to use rating scales for symptom monitoring and titration trials to determine optimal medication dosages. Physicians were taught to prescribe 4 different doses of methylphenidate during a titration trial (placebo, 18 mg, 36 mg, 54 mg); the order of week-long dosing was blinded but standardized across patients (week 1, 18 mg; week 2, placebo; week 3, 36 mg; week 4, 54 mg) to determine optimal dosing for each patient. Parents and teachers completed weekly behavioral ratings (Conners Global Index) & side effect rating scales. Data were returned to Duke Univ psychiatrist to determine the best starting medication dose; a report describing the titration results was faxed back to pediatricians.

Patients in control group practices received treatment as usual, without access to consultative services.

Assessed Conners Global Index & side effect rating scales.

Monthly follow up with Conners and side effect rating scales for 12 months, sent to Duke U psychiatrists for interpretatin, with recommendations returned to the pediatrician

Use of symptom ratings did not differ significantly by group, nor did the change in symptoms over time. Pediatrician compliance with the collaborative consultation service was poor (pediatricians for 29 of 59 patients in the consultation group received a titration trial and 13/59 participated in monthly medication monitoring). Preliminary secondary analyses indicated that those children whose pediatricians complied with titration had significantly better outcomes compared with those who did not and TAU controls (group × time P<.01) Children in the collaborative consultation service–complier group had a 27% reduction in symptom scores compared with 18% reduction in the TAU controls and 13% reduction in consultation non-compliers.

Epstein, 2016²⁵⁵

United States

Cluster randomized controlled trial of either a technology-assisted quality improvement (QI) intervention or TAU control. QI intervention consisted of 4 training sessions, office flow modification, guided QI, and an ADHD Internet portal to assist with treatment monitoring versus TAU control practices

Assessed intervention effects on parent- and teacher-rated ADHD severity using on the Vanderbilt ADHD total symptom score.

12 months follow up

Intent-to-treat analyses examining outcomes (parent ratings of ADHD severity) in all 577 children assessed for ADHD were not significant (b=−1.97, P=0.08), but among the 373 children prescribed ADHD medication, a significant intervention effect on reducing parent-rated symptom severity (b=−2.42, P=0.04) but not teacher-rated symptoms was observed. Prescriber compliance with treatment guidelines was poor, as only 373 of the 577 patients received medication at any time in the 1-year follow-up, and many who did receive it were prescribed sub-optimal doses. Compared with the usual care group, providers in the intervention group had 25% more patient contacts (d=.38, p=.0008) and collected 4.6 (d=.57, p<.0001) and 9.9 (d=.54, p<.0001) times more parent and teacher ratings, respectively. However, providers in the intervention group collected parent ratings in only half and teacher ratings in a quarter of their patients during the initial year of medication treatment.

Fiks, 2017²⁶⁸

United States

Cluster-randomized open label trial at the practice level (9 intervention, 10 control sites) for 3-component quality-improvement program that employs distance learning: (1) 3 15-minute Web-based presentations on evidence-based practices for managing ADHD in primary care; (2) optional collaborative consultation with ADHD experts via a health system online networking site or private email/telephone conversation; (3) and performance feedback reports or calls every 2 months informing them of their rates of sending and receiving ADHD rating scales from parents and teachers and allowed them to compare their results to results of the entire group; feedback reports were discussed during four, 1-hour conference calls). Participation qualified for Maintenance of Certification credit from the American Board of Pediatrics. Collection of rating scales was facilitated via an electronic application linked to the electronic health record versus waitlist control

Number of parent and teacher rating scales sent out and received back assessed

Differences between intervention arms were not statistically significant, though clinicians in both study arms were significantly more likely to administer and receive parent and teacher rating scales compared to an 8-month baseline period. Intervention clinicians who participated in at least one performance feedback call were more likely to send out parent rating scales than intervention clinicians who did not participate (relative difference of 14.2 percentage points, 95% CI: 0.6, 27.7. For all study outcomes, practices with the highest rates of clinician participation in the study (≥ 80%), were not superior to practices with lower rates of involvement (< 80%). Participation was low (105 of 166 invited); 42 of 53 in the intervention group completed all 3 education presentations; 30 (57%) participated in at least one feedback call, and 19 (36%) participated in all 3 components of the intervention.

Florida International University, 2010²⁷⁴

United States

Randomized to receive either osmotic release oral system-methylphenidate alone (78%) or behavioral therapy alone (22%). After 6 months, children with a decline in body mass index >0.5 z-units were randomized to 1 of 3 weight recovery treatments: (1) monthly height/weight monitoring plus daily medication; (2) drug holidays on non-school days (with monthly monitoring); or (3) daily caloric supplements (with daily medication and monthly monitoring).

Standardized body weight and height assessed

18 follow-up visits over 30 months

All groups significantly increased their weight gain. Drug holidays + monitoring, caloric supplementation + monitoring, and monitoring alone all led to increased weight velocity in children taking CNS stimulants, but with no differences between groups, and no intervention led to increased height velocity. When analyzed by what parents did (versus what they were assigned to), caloric supplementation (p<0.01) and drug holidays (p<0.05) increased weight velocity more than monitoring of height and weight. Over the entire study, participants declined in standardized weight (−0.44 z-units) and height (−0.20 z-units).

Oppenheimer, 2019⁴⁶⁶

United States

Naturalistic study of a Web-based platform enabling clinicians to administer online monthly clinical questionnaires to parents and teachers for monitoring of patients remotely between visits. Trigger algorithm alerts clinicians to clinically actionable events that are documented in the medical record versus non-alert group

Patients were the unit of analysis. Parent and teacher reports of current medication, medication side effects inventory, Vanderbilt ADHD Parent Rating Scale, Clinical Global Impression-Severity (CGI-S) scale, and Clinical Global Impression-Improvement (CGI-I) scale

15 months follow up

Trigger algorithms produced alerts requiring immediate review in 8% of the parent reports. Clinicians perceived 74% of alerts to be significant enough to prompt urgent follow-up with parents, suggesting a low rate of false positive alerts. Patients who generated alerts compared to those who did not had more severe ADHD symptoms (beta = 5.8, 95% CI: 3.5–8.1 [p < 0.001] in the 90 days prior to an alert, further supporting validity of the alerts.

Smith, 2000⁵⁴⁵

United States

A assessed the reliability, validity, and unique contributions of self-reports by adolescents receiving treatment for ADHD in a summer treatment program that included self-monitoring as a treatment component.

Self-reported IOWA Conners Inattention/Overactivity and Oppositional/Defiant subscales, ratings of interactions with peers and staff. Assessed changes in reliability during a placebo-controlled, cross-over study of 30 mg of methylphenidate.

Observed frequencies of negative behavior, rating from parents and teachers

Average reliability for the adolescent self-report across all measures was .78 (range .74-.83), similar to the reliability of .82 for counselors (range .78-.85), and significantly better than the teacher reliability of .60 (range .51-.68). Teacher and counselor ratings on the Conners changed significantly during stimulant treatment whereas adolescent self-ratings did not. The findings suggest that adolescents can provide reliable information on their symptoms, but not beyond what parents can provide. Adolescents may also be poor sources of information about the change in ADHD symptoms, but a good source of intormation about improved interactions with others in response to treatment.

Weisman, 2018⁶⁰⁹

Israel

Mobile app allows patients or their parents to report their clinical status following initiation of prescription or after changing medication dosage; purpose of the app is to facilitate communication with MD; app includes questions on severity of ADHD symptoms and potential side effects and can also function as a medication reminder

Treatment as usual, without app

CGI-Severity no significant difference

No significant difference on ADHD-RS, possibly due to inadequate power, Significant difference (p= 0 .008) favoring intervention group on the Clinician Rating Scale (CRS). Intervention group had significantly better adherence, as measured by pill count (p < .015).

Yang, 2012⁶²⁹

Korea

Naturalistic study of symptom monitoring and medication adherence assessed using the Medication Event Monitoring System, a bottle cap with a microprocessor that records all instances and times that the bottle is opened

Patient self-report, clinician rating, pill count assessed; measure of adherence

8 weeks follow up

The rate of non-adherence was 46.2%, higher than patient self-report of 17.9%, clinician rating of 31.7%, and pill count of 12.8%. Pill count and monitoring system concordance was 0.249 (95% CI: 0.102-0.386). Self-report concordance was 0.237 (95% CI: −0.024-0.468). Non-adherent patients had more severe symptoms at baseline and inferior improvement compared with adherent patients.

: Notes: ADD-H = attention deficit disorder with hyperactivity, ADHD = attention deficit hyperactivity disorder, ADHD-RS = ADH rating scale, CI = 95% confidence interval, CGI-I = Clinical Global Impression-Improvement scale, CNS = central nervous system stimulants, CRS = clinician rating scale, GGI-S = (Clinical Global Impression-Severity scale, KQ = Key Question, MD = medical doctor, QB test = quantified behavioral test, RR = relative risk, TAU = treatment as usual

From: 6, Results: Monitoring ADHD

Cover of ADHD Diagnosis and Treatment in Children and Adolescents

ADHD Diagnosis and Treatment in Children and Adolescents [Internet].

Comparative Effectiveness Review, No. 267.

Peterson BS, Trampush J, Maglione M, et al.

Rockville (MD): Agency for Healthcare Research and Quality (US); 2024 Mar.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.