Developmental disorder hospitalizations in Canada before and during the COVID-19 pandemic: Technical notes

---

Data source

Data used in this study are from the Canadian Institute for Health Information's (CIHI) Discharge Abstract Database (DAD). The DAD contains demographic, administrative and clinical information on hospital discharges from acute inpatient facilities in all provinces and territories, except Quebec.

---

Study sample

The study sample includes all hospitalizations in Canada (except Quebec) between fiscal years 2015/16 and 2021/22. For the purpose of this analysis, the study sample was divided into the following groups:

• Developmental disorder (DD) hospitalizations: patients with a DD code in any diagnosis field (n=141,001)
• All hospitalizations: patients with any diagnosis, including DD (n=17,568,081)

Developmental disorder hospitalizations were identified in the DAD using the relevant International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Canada (ICD-10-CA) codes (see Table A for a detailed list of included codes). The group of conditions selected were chosen through consultations with academic clinicians who work in the DD field.

Where relevant, DD hospitalizations were compared to all hospitalizations. This includes any hospitalization captured in the DAD, regardless of the ICD-10-CA codes associated with the hospital discharge.

All acute care hospitalizations between 2015/16 and 2021/22 were included in the analysis except:

• residents of Quebec
• non-residents of Canada
• stillbirths or cadaveric donors
• where sex was recorded as other than male or female
• where age was invalid (age missing or greater than 114 years)

These groups were excluded in order to align the population included in numerator and the denominator in all rate calculations. Additionally, small cell sizes where sex was recorded as other than male or female prevented their inclusion in the analysis.

Table A. Developmental disorder case definition

Developmental disorder category	ICD-10-CA codes
Mental and behavioural disorders	F70.x Mild mental retardation F71.x Moderate mental retardation F72.x Severe mental retardation F73.x Profound mental retardation F79.x Unspecified mental retardation F78.x Other mental retardation F80.0 Specific speech articulation disorder F80.1 Expressive language disorder F80.2 Receptive language disorder F80.8 Other developmental disorders of speech and language F80.9 Developmental disorders of speech and language, unspecified F81.x Specific developmental disorder of scholastic skills F82.x Specific developmental disorder of motor function F83.x Mixed specific developmental disorders F84.x Pervasive developmental disorders F88.x Other disorders of psychological development F89.x Unspecified disorder of psychological development F90.x Hyperkinetic disorders F95.x Tic disorders F98.4 Stereotyped movement disorders F98.5 Stuttering F98.6 Cluttering
Diseases of the nervous system	G80.x Cerebral palsy
Certain conditions originating in the perinatal period	P04.x Fetus and newborn affected by noxious influences transmitted via placenta or breast milk
Congenital malformations, deformation and chromosomal abnormalities	Q05.x Spina bifida Q86.x Congenital malformation syndromes due to known exogenous causes, not elsewhere classified Q87.0, Q87.1, Q87.3, Q87.5 Other specified congenital malformation syndromes affecting multiple systems Q90.x Down syndrome Q91.x Edward's syndrome and Patau's syndrome Q93.x Monosomies and deletions from the autosomes, not elsewhere classified Q99.2 Fragile X chromosome

---

Variables

• Year: The fiscal year of hospital discharge. Each fiscal year begins on April 1. For example, 2020/21 indicates the patient was discharged between April 1, 2020, and March 31, 2021.
• Sex: The sex of the patient, defined as male or female.
• Age: The age of the patient at time of hospital admission, categorized into age groups (0-6, 7-11, 12-18, over 18 years).
• Month: The month of hospital discharge.
• Reason for hospitalization: The reason for hospitalization is the condition most responsible for the patient's hospital stay. In the DAD, this condition was assigned an ICD code.
  • These ICD codes were analyzed individually (by disorder) as well as by ICD disease chapter (by disorder group). More detailed information about each ICD chapter and individual codes is available from CIHI.
  • COVID-19 codes were as follows (more info at CIHI: COVID-19 data collection and coding direction):
    • U07.1 COVID-19, virus identified (confirmed COVID-19)
    • U07.2 COVID-19, virus not identified (suspected COVID-19)
  • COVID-related codes were as follows:
    • U07.3 Multisystem inflammatory syndrome associated with COVID-19
    • U07.4 Post COVID-19 condition
    • U07.5 Personal history of COVID-19
    • U07.6 Need for immunization against COVID-19
    • U07.7 COVID-19 vaccines causing adverse effects in therapeutic use
• Length of stay: The total length of stay is the total number of days the patient was hospitalized. This was derived in the DAD by subtracting a patient's admission date from their discharge date. Total length of stay is a combination of the acute length of stay (the number of days the patient was in acute care) and the alternate level of care (ALC) length of stay (the number of days the patient was occupying a bed in a facility and did not require the intensity of resources or services provided in that care setting). The mean (average) total length of stay was used in this study.

---

Methods

Statistical analysis

• A descriptive analysis of hospitalizations in Canada (except Quebec) between fiscal years 2015/16 and 2021/22 using data from the DAD was carried out.
  • The proportion of all hospitalizations that were DD hospitalizations was examined, in addition to the distribution of sex and age among DD and all hospitalizations.
  • Trends in overall, sex-specific and monthly hospitalization rates (crude and age-adjusted), and age-specific hospitalization rates were examined by year, for DD hospitalizations and all hospitalizations. Statistic Canada's 2011 Canadian Census of Population was used to age-adjust hospitalization rates using 5-year age groups.
    • Note: Hospitalizations with an unknown discharge month (n=20) were included in all analyses except for the monthly rates.
  • Descriptive analysis was conducted to explore the reasons for hospitalization among DD hospitalizations.
  • Trends in the average length of stay of both DD hospitalizations and all hospitalizations were examined.
    • A sensitivity analysis was conducted as part of the length of stay analysis. It explored the impact of an outlier with an extremely long length of stay among the 2020/21 DD hospitalizations. This outlier was removed from the sample and no significant differences in the findings were observed. Therefore the outlier was retained in the sample for all analyses.
• 95% confidence intervals for rates were calculated using the following formula:

  95% confidence interval = rate +/- [1.96(sqrt(variance))]
  where:
  variance = (number of hospitalizations/(Canadian population estimate^2))

• Statistical differences between 2 estimates were determined by comparing the associated 95% confidence intervals. Non-overlapping 95% confidence intervals between estimates were considered statistically significantly different.
• Analyses were conducted using SAS Enterprise Guide, version 7.12 (SAS Institute Inc., Cary, NC, USA).

Proportion and rate calculations

• Proportion of all hospitalizations that are DD hospitalizations was determined by dividing the number of DD hospitalizations by the total number of hospitalizations for that fiscal year, then multiplying by 100.
• Rate of DD hospitalizations (per 100,000) was determined by dividing the number of DD hospitalizations by the Canadian population (except Quebec) for that fiscal year, then multiplying by 100,000. A similar approach was used to calculate the sex-specific, age-specific, and monthly rate of DD hospitalizations.
• Rate of all hospitalizations (per 100,000) was determined by dividing the total number of hospitalizations by the Canadian population (except Quebec) for that fiscal year, then multiplying by 100,000. A similar approach was used to calculate the sex-specific, age-specific, and monthly rate of all hospitalizations.
• Percent change in rate compared to the previous year was calculated by subtracting the rate in the current year from the rate in the previous year, then dividing by the rate in the previous year and multiplying by 100.
• Proportion of DD hospitalizations with COVID-19 (confirmed or suspected) while in hospital was determined by dividing the number of DD hospitalizations with a confirmed or suspected COVID-19 ICD code (U07.1 or U07.2) in any diagnosis field by the total number of DD hospitalizations for that fiscal year, then multiplying by 100. Due to the small number of COVID-19 cases in 2019/20, this was only calculated for the 2020/21 and 2021/22 fiscal years. A similar approach was also used to calculate the proportion of confirmed or suspected COVID-19 for all hospitalizations.

---

Limitations

• The capture of a DD hospitalization depends on a DD code being recorded in the hospital discharge abstract. For this reason, DD hospitalizations within this study:
  • do not necessarily represent a confirmed DD (could represent a suspected or unconfirmed condition)
  • may not represent all individuals with DD who are admitted to hospital, since a DD code may not have been recorded in the abstract, especially those hospitalized for another medical reason
• The data presented here only represent acute care hospitalizations. Therefore, they don't represent the full health care picture during the pandemic period.
• The DAD doesn't contain information on hospitalizations from Quebec. Therefore, a true national analysis of hospitalizations during the pandemic isn't possible using this data source.

---

Glossary

Age-specific rate: Rate calculated for a specific age group.

Age-adjusted (or age-standardized) rate: A measure of the rate that a study population would have if it had the same age structure as the standard population. Age-adjusted rates are commonly used when comparing estimates over time (trend analysis) or when comparing rates for different geographical areas or subpopulations.

Crude rate: Rate determined by dividing the total number of cases in a given time period by the total number of persons in the population.

Confidence interval: A statistical measurement of the reliability of an estimate. The size of the confidence interval relates to the precision of the estimate. Narrow confidence intervals indicate greater precision than wider ones. The 95% confidence interval shows an estimated range of values that's likely to include the true value 19 times out of 20.

Sex-specific rate: Rate calculated for a specific sex (male or female).

---