Because understanding and producing a language are different skills, and many children get enormous listening practice at home while getting almost no chance to speak Vietnamese with peers. Linguists call this receptive bilingualism, and it is one of the most common outcomes in immigrant families.
The typical pattern looks like this. Parents speak Vietnamese at home. The child answers in English. The parents understand English well enough that the conversation works, so nobody corrects it. The child keeps absorbing Vietnamese input and stops producing Vietnamese output — and speaking is the skill that atrophies.
School accelerates it. Once formal education, friendships, television, and homework are all in English, English becomes the language the child thinks in, and Vietnamese becomes the language they only hear.
Vietnamese adds a specific difficulty: tones and the kinship pronoun system. A speaker who is unsure of the tone or the right title will often opt out rather than risk an error in front of relatives who will notice. Embarrassment does more damage to spoken fluency than lack of vocabulary does.
The encouraging part is that receptive knowledge is real knowledge. Adults who "only understand" typically regain speaking ability far faster than someone learning from zero, because the sound system and much of the vocabulary are already in place.
Extremely common across immigrant communities, but not universal — outcomes vary with how much Vietnamese-speaking community a child grows up around, whether grandparents live nearby, and whether the family attends Vietnamese language school.