Overview
An analysis of cause-of-death data from 2013 to 2022 reveals a noteworthy shift in cancer mortality among New Zealand children aged 0 to 4, with a specific emphasis on infants under one year of age. The findings are derived from the Ministry of Health’s ICD-10 coded mortality dataset (Chapter 2: Neoplasms), filtered for the age groups “0” (infants under 12 months) and “00-04” (all children from birth to age 4).
Key Findings
| Year | Age 0 Deaths | Age 00–04 Deaths | % of 00–04 that were Age 0 |
|---|---|---|---|
| 2013 | 96 | 232 | 41% |
| 2014 | 64 | 192 | 33% |
| 2015 | 112 | 280 | 40% |
| 2016 | 40 | 136 | 29% |
| 2017 | 16 | 176 | 9% |
| 2018 | 64 | 184 | 35% |
| 2019 | 56 | 144 | 39% |
| 2020 | 48 | 184 | 26% |
| 2021 | 48 | 312 | 15% |
| 2022 | 72 | 168 | 43% ← unusually high |
- Record High Proportion of Infant Cancer Deaths in 2022
- In 2022, there were 72 cancer deaths among children aged 0 (infants), representing 43% of all cancer deaths in the 00-04 age band.
- This is the highest proportion of infant cancer deaths in the full decade of available data (2013–2022).
- Significant Drop in Overall 00-04 Cancer Deaths Post-2021 Spike
- Cancer deaths in the broader 00-04 group peaked in 2021 at 312 deaths, then sharply declined to 168 in 2022.
- However, despite the overall drop, infant deaths rose from 48 in 2021 to 72 in 2022.
- Long-Term Trends in Age 0 vs 00-04
- From 2013–2020, the average number of cancer deaths in age 0 was approximately 62.3 per year.
- In 2022, the number is above average, and it occurred during a year where the rest of the 00–04 group saw a substantial decline.
- Temporal Alignment with Vaccine Uptake
- The 2021–2022 period overlaps with the mass rollout of COVID-19 vaccination in adults, including pregnant women.
- Infants born in 2022 are likely to have been in utero during the peak vaccination period of 2021.
- No causation is implied here, but the temporal alignment is noted for further investigation.
Expanded Context – All-Cause Mortality Changes
A comparison of total mortality across all causes (ICD Chapters) in the same dataset revealed:
| ICD Subgroup (0-0) | Avg Deaths (2013–2019) | Deaths in 2022 | Change vs Baseline |
|---|---|---|---|
| C15–C26: Digestive organ cancers | 0.0 | 12 | +12.0 |
| D10–D36: Benign neoplasms | 10.4 | 12 | +1.6 |
| All_ICDSubgroups (total) | 32.6 | 36 | +3.4 |
| D37–D48: Neoplasms of uncertain/unknown behavior | 12.7 | 12 | ~–0.7 |
| C73–C75: Thyroid and related glands | 14.0 | 0 | –14.0 |
- Large increases in deaths from perinatal conditions (ICD Chapter 16) and ill-defined symptoms or signs (ICD Chapter 18).
- Cancer deaths (Chapter 2) and special-purpose coding (Chapter 22) also showed marked increases post-2020.
In 2022, 12 infant deaths (age 0) were coded under ICD Chapter 22: Special Purpose Codes, all within the U07–U10 range, which includes COVID-19-related classifications. This is a marked rise from zero in prior years. The appearance of even 12 such cases raises concern, as publicly reported COVID-19 fatalities in this age group were negligible — and total COVID-coded deaths in the entire 0–10 cohort did not exceed this figure across all available reporting sources. This suggests either:
A shift in certification practice (e.g., inclusion of incidental positive tests),
Or the coding of neonatal or perinatal deaths in COVID-exposed pregnancies under Chapter 22, despite no direct causal confirmation.
While these cases may not represent deaths from COVID, their presence within the cause-of-death registry should be more transparently accounted for.
Visual Trends
Two line charts track annual mortality counts from 2013 to 2022 for the four ICD chapters with the greatest excess in:
- Infants (Age 0): Steady rise in perinatal and ill-defined categories; cancer and special code entries increase post-2020.
- Children (Age 00–04): Similar trend, with a pronounced 2021 cancer spike before falling in 2022.
Additional analysis of neoplasm subgroups revealed that in infants, the entire 2022 increase in cancer deaths was attributable to the sudden appearance of digestive organ cancers (C15–C26), a subgroup that was absent in prior years.
Simultaneously, thyroid and endocrine gland cancers (C73–C75)—previously present—disappeared completely in 2022. No such pattern occurred in ages 1–4.
This suggests the anomaly is exclusive to infants and not likely to be a result of broad coding changes across the 0–4 age band. Visual plots confirmed the substitution-like trend in infants, while ages 1–4 remained flat for both subgroups.
Next Steps
- Analyze whether this pattern persists into 2023 once data becomes available.
- Conduct normalization by birth cohort size to determine if the mortality rate per 1,000 births also increased.
- Examine ICD subgroup breakdowns for cancer types contributing to the infant death spike.
- Further investigate causes behind the rise in deaths recorded under Chapter 18 and Chapter 22.
- Consider whether internal MoH or DHB coding guidance may have shifted classification practices mid-decade.
Data Source: Ministry of Health NZ, ICD-10 Chapter 2 and full-cause mortality data (2013–2022).
Were More Babies Born?
To test whether these increases might simply reflect a growing infant population, we examined annual average birth cohort sizes (proxied by age 0 population):
| Year | Average Age 0 Population |
| 2013 | 59,858 |
| 2014 | 58,370 ⬇️ |
| 2015 | 59,018 |
| 2016 | 59,718 |
| 2017 | 59,958 |
| 2018 | 59,408 |
| 2019 | 59,808 |
| 2020 | 59,238 |
| 2021 | 61,060 ⬆️ |
| 2022 | 59,855 |
The 2021 birth cohort was modestly larger than surrounding years, but not enough to explain the sharp post-2020 increases in infant deaths. 2020 and 2022 were typical and align with pre-pandemic norms.
