**Background:** Globally, 2.5 million newborns die annually before age one month, with most deaths in low- and middle-income countries. Inadequate knowledge and skills for newborn care contribute to these deaths. The Care Companion Program (CCP) is a hospital-based family caregiver education program implemented in six Indian states that teaches key preventive behaviors including exclusive breastfeeding, dry umbilical cord care, handwashing, skin-to-skin contact, nonrestrictive maternal diet, and danger sign recognition. This qualitative study aimed to understand how and why knowledge and behaviors change as a result of CCP sessions and participants' perceptions of the program's impact.
**Methods:** The study was conducted in four district hospitals in Karnataka, India (Chikkaballapur, Ramanagara, Dharwad, and Bagalkot). Forty-two in-depth interviews were conducted among families with recently-delivered babies (23 mothers and 19 family caregivers) from different families. Participants were purposively recruited from a larger ongoing study based on whether newborn complications were reported (23 reported complications, 19 did not). Interviews were conducted in participants' homes from February-March 2020 by a trained third-party external agency, audio-recorded, transcribed, and translated into English. A codebook was developed using inductive and deductive coding with a doer/non-doer analysis framework. Interviews lasted between 20–80 minutes.
**Key Results:** Overall, 30 respondents found the training useful and 32 would recommend CCP to a friend. Nineteen mothers said having family members present during CCP made them feel supported. For exclusive breastfeeding, all 42 respondents reported feeding breastmilk, but several supplemented with cow's milk (n=5), glucose water (n=1), gripe water (n=7), honey (n=2), and other substances. The most commonly mentioned barrier was lack of milk production. For handwashing, 35 respondents practiced regular handwashing before CCP, and 14 attributed to CCP that they now regularly wash with soap before touching their baby. For skin-to-skin care, 33 learned about it from CCP, but 25 reported not practicing it, and understanding of how and when to perform it varied considerably. For cord care, 37 did not put anything on the cord, while 5 applied substances (hot coconut oil, powder, castor oil, boric powder) before it fell off. For maternal diet, 38 restricted specific foods/liquids postpartum despite 33 recalling that CCP recommended no restrictions. Common restrictions included foods believed to cause cold, "hard foods," spicy foods, and water (n=8). Only 4 respondents reported no restrictions, with 3 directly attributing this to CCP. For danger sign recognition, 4 respondents said CCP influenced their knowledge of danger signs, 4 said it influenced symptom management, and 3 said it influenced health-seeking behavior.
**Clinical Implications:** The CCP program positively influenced key neonatal behaviors by imparting knowledge and skills, but the effect was not universal across all health behaviors. Behaviors that were already culturally accepted (handwashing) or where CCP provided new actionable details (breastfeeding technique) showed more change, while behaviors conflicting with strong cultural norms (maternal diet restriction) showed limited change. The study highlights the importance of including family members, especially grandmothers who often control maternal diet, in education programs. The findings suggest opportunities for program improvement including greater consistency across trainers, directly addressing socio-cultural norms that conflict with medical advice, and allowing for more interactive question-answer sessions. These results complement a previous pilot evaluation of CCP that reported statistically significant improvements in dry cord care practice (4%), skin-to-skin care (78%), newborn complications reduced by 16%, mother complications by 12%, and newborn readmissions by 56%.