Codeswitching in the context of single or multiple conversations has been a myth for language experts. Matrix language framework (MLF) model proposed by Myers Scotton (1993) has become very popular for the analysis of language pairs, and is influential in determining matrix language in different language pairs. The aim of this study is to identify matrix and embedded language in Urdu-English data sets of health and science theme. MLF model is applied to an original article on Covid-19. Data sets include language pairs from a published article on the nature of coronavirus. A qualitative design was followed to arrange data sets, language pairs were identified, transcribed and coded carefully according to the Canonical Trilinear Representation. Three layers of data with the first layer of roman Urdu, the second layer of gloss and the third layer of English translation were further analyzed syntactically and morphosyntactically to show how they grammatically occur in the bilingual complementiser phrases. The findings of this study reveal that code-switching was permissible even when it led to structural asymmetry. English insertions received different Urdu markers of gender and number wherever required. Urdu adjectives played a significant role in realizing nouns. Some data sets allowed English insertions without Urdu markers. Moreover, the data supported matrix language frame, morpheme order principle and system morpheme principle and no counter example appeared against MLF model. Thus, the present study is a significant contribution in the related area.