Tables 7 and 8 assume that mortality rates in 1911 and 1921 were basically the same, or at least that any changes were too small to be discerned in statistics as approximate as those in Tables 7 and 8. In fact, major improvements in the health of the New Territories only began to be seen in the 1930s. The government had begun to provide for the health of the New Territories immediately after the take-over. A dispensary with a resident doctor was established at Tai Po in 1900, and a programme of vaccination was initiated. However, most of this clinic's time was taken up with treating the police and other government servants in the New Territories; few local villagers were affected. The annual number of vaccinations (including re-vaccinations) was under 100 before 1909, when the Tung Wah Hospital began to assist every summer. Between 1910 and 1920 the annual vaccination rate in the New Territories rose to several hundred. Nonetheless, the area affected by this vaccination campaign seems to have been limited to the market towns of Tai Po, Sheung Shui and Tsuen Wan. The doctor posted to the New Territories in 1900 was withdrawn in 1909, after which the only trained medical staff resident in the area was a “dresser”, who was supervised by intermittent visits by a doctor from Kowloon.
The very high rates of neo-natal casualties in the New Territories began to be addressed by the government from 1914, when a government midwife was stationed at Yuen Long. Midwives were posted to Tai Po, Tsuen Wan, and Cheung Chau in 1915, 1916, and 1917 respectively. The Pok Oi Hospital at Yuen Long was reorganised in 1920, and was backed by a government dispensary from 1925. However, the critical decision to post midwives to cover the villages in addition to the market towns was only taken in the 1930s. From the 1930s, the district midwives disinfected all drinking water wells, and vaccinated against smallpox, as well as attending births. Oral evidence suggests that the results on infant mortality were massive. By 1921, however, these great improvements had only begun to affect the market towns, and the overall mortality rates for the New Territories as a whole must have been much as they had been in 1911 and earlier. Between 1911 and 1921, changes in mortality rates were probably, therefore, marginal, and averaging the two sets of statistics, as in Tables 7 and 8, is not unreasonable.