Data analysis uncovered three main themes: (1) Antibiotic prescribing decision is a multifactorial process influenced by environmental and sociocultural factors; (2) IPC, diagnostics and treatment need strengthening; (3) AMR is a social problem that requires a collective effort. The analysis process that led to the creation of (sub)themes from (sub)categories, codes and meaning units is presented in the coding tree (Table 2). Themes and subthemes were mapped against the COM-B-TDF framework (Fig. 1).

Theme 1- antibiotic prescribing decision is a multifactorial process influenced by environmental and sociocultural factors

General and orthopedic surgeons indicated that their antibiotic prescribing decisions were largely guided by medical knowledge and clinical assessment of factors such as the type and severity of injury, type of surgery and signs of infection. Interviewees felt they made good decisions about antibiotic prescribing based on patient follow-up, when it showed improvement in the patient’s condition. They also emphasized the critical role of antibiotics in surgical care, expressing views like, “antibiotics are most important in orthopedic surgeries” (O-02) and “surgical management is all about injectable antibiotics” (S-12). Furthermore, special considerations that affect their clinical decision-making on antibiotic prescribing were revealed, including the desire to achieve positive patient outcomes despite diagnostic uncertainty and the specific challenges of treating orthopedic infections. Orthopedic surgeons noted that treating complications in orthopedics is often a long and painful process with uncertain outcomes.

Antibiotic prescribing decisions are influenced by environmental and sociocultural factors

Throughout the interviews, context-specific environmental and sociocultural factors influencing antibiotic prescribing emerged as a major topic. It was often mentioned that dry and dusty rural environment “pushes the use of antibiotics” (S-08). Participants felt that many more factors influence clinical decisions on antibiotic prescribing in India compared to high-income countries, such as a high patient load and excessive workload for doctors, the quality of medical education and available facilities, and the influence of pharmaceutical companies on doctors’ and patients’ behavior. Furthermore, interviewees highlighted patients’ socioeconomic status as a critical factor affecting antibiotic prescribing and patient recovery in India. It was frequently mentioned that patients in the area were poor; therefore, the affordability of antibiotics was a major factor in treatment decisions, along with the costs of diagnostics and surgeries. Furthermore, it was often noted that a patient’s compliance and trust in the doctor depended on the affordability of the prescribed antibiotics. Consequently, doctors often needed to modify their prescriptions, sometimes opting for simpler, less expensive antibiotics even when patients needed “higher” broad-spectrum antibiotics, which can be very costly.

“If we have to switch to the higher antibiotics like the piperacillin-tazobactam, the combination ones, two-three antibiotics at a time…So, these are the very costly ones, most patients cannot afford. Instead of that, they choose to quit the hospital…You just give whatever you can. So, we have to switch to a lower spectrum antibiotic. Even if it’s not working. Because sometimes you know this antibiotic [is] resistant, even then we are prescribing it because we don’t have any other option. We cannot go to get colistin for every patient. The last option left for all resistant patients is colistin. But one injection costs around 1000 bucks. And their monthly income is 1000 rupees. So, they cannot afford it. So, we have to give something, even if it is resistant. So at least for moral grounds, we are treating the patient. Otherwise, it is just that the patient is left on their own.” (S-13).

Further significant factors that influence clinical decision-making on antibiotic use were patient education and behavioral factors: illiteracy, poor hygiene practices, late presentation, loss to follow-up, low treatment compliance, a desire to get better instantly, and patients’ pressure and beliefs about antibiotics. For example, patients believed that the treatment without antibiotics was incomplete and consequently, doctors were perceived better if they prescribed more antibiotics and steroids. In general, “the more antibiotics, the better” (O-05) was a wide-spread patient belief.

“…and in private practice, if patient is not responding very well, means no response to our treatment, patient will switch over to other fellow, other doctor… So, they want miracles. Within a week they want to be cured.” (S-09).

Decision on the type and duration of antibiotic treatment is multifactorial

The choice of antibiotic was found to be influenced by multiple factors, including the patient’s condition and history, the antibiotic’s pharmacokinetic and pharmacodynamic properties, the prescriber’s experience, and personal preference. Prescriber experience was particularly emphasized, with many interviewees expressing that they relied on their knowledge of the most common organisms causing infections to guide their antibiotic selection. Similarly, personal preferences in prescribing were highlighted, as participants frequently mentioned that different doctors had different prescribing practices, such as preference of starting with oral antibiotics, using only broad-spectrum antibiotics, or using antibiotics only for operative cases, etc.

“The only thing is, few antibiotics have been inoculated in our culture, in our prescription culture, because of our practices, which we have seen, like which our seniors have seen through their experiences, that this antibiotic is working well in such scenario.” (S-12).

Surgeons generally favored broad-spectrum antibiotics due to their effectiveness and ability to cover a wide range of bacteria. The treatment was most frequently initiated with broad-spectrum antibiotics to control and prevent the spread of infection because clinicians “cannot individualize antibiotic therapy from day one” (O-05). The use of broad-spectrum antibiotics was considered standard practice in the surgical field, with many interviewees having positive experiences with broad-spectrum antibiotics in their setup.

“Otherwise, from my experience of last 23 years, I have operated so many cases of perforation, obstruction, but this is my experience that I can say proudly that I have saved most of the patients. And definitely, there is a role of this broad-spectrum antibiotics.” (S-08).

The most emphasized benefit of narrow-spectrum antibiotics was their ability to provide targeted treatment. However, participants rarely started the treatment with narrow-spectrum antibiotics, except in cases of minor injuries or surgeries, due to uncertainty about their effectiveness. When asked about how often they switched from broad-spectrum to narrow-spectrum antibiotics, responses varied widely. Some reported doing so in most cases, while others stated it happened only in very few cases, as they typically did not switch to narrow-spectrum if broad-spectrum antibiotic was effective. Reasons for switching from broad-spectrum to narrow-spectrum antibiotics were similarly inconsistent. Some interviewees opted to switch when the patient’s condition was improving, while others did so when the condition was not improving, guided by culture and susceptibility reports.

Surgeons reported a wide range of antibiotics they commonly prescribed, with amoxiclav, aminoglycosides and third-generation cephalosporins being the preferred choices. They generally started treatment with third-generation cephalosporins, with the addition of amikacin and metronidazole for contaminated wounds or major injuries.

In terms of decisions on the duration of antibiotic treatment, interviewees considered the following factors: the type and severity of the injury, IPC measures, and the type of surgery. The duration of perioperative surgical prophylaxis was not standardized and ranged from one to three doses. Many surgeons acknowledged that determining the duration of antibiotic treatment was problematic and not defined. Intravenous broad-spectrum antibiotics were generally given for two to five days, with extended durations in major surgeries and severe cases, particularly in certain orthopedic conditions (e.g., osteomyelitis and septic arthritis). Oral antibiotics were commonly prescribed for two to three days during hospitalization and five to seven days after discharge. Some participants revealed that they prescribed antibiotics for longer durations due to fear of postoperative infections or patients not returning for follow-up.

“Sometimes for the patient, for our own mental peace also, we step up the antibiotic so that we are avoiding an unforeseen infection, which is not justifiable, because what if that infection never occurred? So, I think we need to have a protocol for this thing also…So if someone tells us or some kind of proof comes that our antibiotics, even oral antibiotics, can work or if we step down then it won’t be an issue. Then I think that will help us to at least decrease our use in antibiotics for the duration. If someone tells us that even a 3-day duration is enough and we have a proof for that, then we can try that. We need not go for a 5-day duration.” (S-12).

The need for feedback and consultation on antibiotic prescribing depends on seniority level

When asked whether they followed any guidelines for antibiotic prescribing and which ones, some participants stated that they adhered to international and national guidelines, as well as a “general standard protocol” or “departmental and hospital guidelines.” However, most participants indicated that there were no formal guidelines for antibiotic prescribing in India, nor written departmental or hospital guidelines, stating that “there cannot be universal guidelines for every patient in every part of the world” (O-05).

There was considerable variation in responses regarding feedback on antibiotic prescribing. Some interviewees reported not receiving regular feedback, while others indicated they received feedback primarily from seniors, colleagues, and patients. The majority of participants viewed regular feedback as valuable for optimizing antibiotic treatment and expressed openness to regular prescription audits.

“It [feedback]would be helpful for us because now, since we are a team, we have a different mode of treatment, different types of patients. So, we can generalize this and then we can get into practice, whichever is most favorable among them [antibiotics]. That would be good for us.” (S-03).

The responses regarding the need to seek advice on antibiotic prescribing were similarly varied. While some senior participants felt confident in their prescribing decisions and seldom, if ever, sought advice, others indicated they consulted microbiology, pharmacology, internal medicine, or pathology departments for complex infectious cases. Most junior residents, however, reported regularly seeking guidance from senior consultants and professors.

“Like our consultants, usually when we are on a round…, they always explain us why we need to change the antibiotic or what is the protocol, why you should change it or which to start and which to stop and for how many days, which is more important.” (S-05).

Theme 2- infection prevention and control, diagnostics and treatment need strengthening

Infection prevention, control and treatment can be improved

Some surgeons pointed out issues with the operating theater conditions and the preoperative preparation of both surgeons and patients. It was also noted that overall cleanliness of the hospitals could be greatly improved, as patients were considered more likely to acquire infections after leaving the operating theater than within it. When reflecting on surgical site infections (SSIs), many participants noted that infection rates were low in both orthopedic and surgical departments. However, when SSIs did occur, they highlighted the severity of the consequences, describing them as a “disaster” and a “failed surgery” (O-06).

Most surgeons saw their antibiotic prescribing practices as good and justified, mentioning that they did not overuse antibiotics and were very restricted on the use of “higher” antibiotics. Several interviewees stated that there was no external pressure or restrictions on the use of antibiotics. Senior surgeons reflected on how antibiotic prescribing has improved over time through education from conferences and symposiums and acknowledged that “20 years ago, we were using much more antibiotics than today” (S-08). Junior surgeons appeared to be more critical of departmental antibiotic prescribing practices and saw room for improvement. It was often mentioned how their prescribing is based on theoretical knowledge from medical books, as well as on what senior doctors had taught them from their experience.

“In our setup, I think the protocols need to be a bit strict…For the broad-spectrum antibiotics, there should be a teaching at least, and also an explanation, like not at the level of junior residents. It should begin at the level of consultant because the one who teaches is the one who should be learned most…Because antibiotics is a thing, the protocol changes. So, though they know everything, but still they need to be updated on this. There should be a program or something that keeps them updated every 3–6 months so that they can keep us updated on these things. Because if you teach a junior resident, they can’t impose things unless their superior is ready for it. So, the things should begin from the top level…The channel should be that way, it should not go the other way. Because residents can never have a say in anything unless the consultants are ready for it. Even justified use of antibiotics should be understood from their levels.” (S-12).

Ambiguous practice of sending samples for culture and susceptibility testing

Most surgeons stated that their standard practice was to send a sample for culture and susceptibility testing before empirically starting broad-spectrum antibiotics, with treatment later adjusted based on the test results. However, there was significant ambiguity regarding whether samples are consistently sent for culture and susceptibility testing, and if so, the exact timing of when the samples are sent remained unclear.

Participants noted they used various clinical criteria to decide whether to send samples for culture and susceptibility testing, though these criteria differed among them. As a result, their responses regarding the frequency of sending different types of samples varied. For instance, the reported frequency of sending blood cultures ranged from “in very few cases” to “in all infected cases.” Similarly, the frequency of sending pus cultures ranged from “rarely” to “very often.”

Most surgeons believed that sending samples for culture and susceptibility testing was very important for ensuring evidence-based, targeted, and appropriate antibiotic use, as well as for investigating local resistance patterns.

“You need to know what are the bugs around and if you know what are the bugs around then you can treat them properly. Otherwise, it is all shot in the dark.” (S-14).

Although many participants were, in general, satisfied with laboratory services in hospitals, they identified several problems related to the practice of sending samples for culture and susceptibility testing. Some interviewees complained about the long waiting time for results, which was generally two to three days but could be delayed up to five days with an increased sample load. In addition, some participants indicated the occasional lack of resources in the departments, e.g. sample tubes and gloves. Several participants revealed that they sometimes could not trust culture and susceptibility reports due to inappropriate handling of samples, such as contamination during collection or transit, or misplacement of samples due to errors in labeling and identification.

Various suggestions were proposed to improve the process of sending samples for culture and susceptibility testing. The need for the development and implementation of a standardized sample collection protocol was emphasized, along with regular training and supervision of staff on proper procedures for collecting, transporting, and storing samples. Similarly, there was a call for clear guidelines on the types of cases and the frequency for mandatory sending of samples, both before initiating and during antibiotic treatment, to ensure appropriate use. Some interviewees suggested expanding the range of antibiotics tested for susceptibility, as it is “no use for a clinician to get the report which is resistant to all antibiotics” (O-05). Additionally, improvements in the notification system from the laboratory to clinicians regarding sample reception and culture results were suggested. It was highlighted that every unit and department should have sufficient stocks of necessary materials, such as sample tubes. Ideally, some surgeons proposed that each department that frequently requires laboratory services should have its own laboratory, and that faster, alternative diagnostic methods should be developed.

Chaotic medical practices and antibiotic prescribing during the COVID-19 pandemic

Many participants evoked the topic of COVID-19 and its impact on their medical practices, particularly in relation to antibiotic prescribing. One senior surgeon noted that during the pandemic, he encountered unusually severe infections where the causative pathogens were unidentified, leading to the highest mortality rates he had ever seen in his practice. Participants recalled the uncertainty they were facing when deciding about treatment protocols, which were frequently changed by the WHO and national health authorities. It was often stressed how antibiotics and steroids were largely misused during pandemic, as broad-spectrum antibiotics were given to every patient, and how people were self-medicating and randomly buying medicines OTC. Additionally, the government distributed “Coronavirus treatment kits” to COVID-19 suspected and confirmed cases. These kits contained antibiotics (azithromycin, doxycycline) and steroids, which people were taking while isolating at home. The overall situation was described as “bizarre-everything was going on” (S-13).

“… and organisms developing a resistance because of mismanagement at the level of antibiotic because which has occurred during COVID. During COVID, no one knew what would work. So, everyone has tried every antibiotic on almost many patients and that has helped the viruses and the bacteria to outgrow such antibiotics.” (S-12).

Theme 3- AMR is a social problem that requires a collective effort

Surgeons shared their observations regarding the development of AMR in their practice, noting that bacteria have become resistant not only to “common basic antibiotics”, but also to broad-spectrum antibiotics, such as third-generation cephalosporins and carbapenems. Most participants said that they were now more aware of the AMR and expressed concern, stating that there is “no assurance anymore that we will be on the safe side when giving a broad-spectrum antibiotic” (O-05). The development of AMR was reflected in the need to change the prescribing protocol every 10 to 15 years. When asked about how often they encounter AMR, most participants indicated that they rarely experienced resistance to all antibiotics, though they acknowledged that the trend of resistance to all antibiotics is increasing. Nonetheless, many reported frequently encountering patients with resistance to some antibiotics, particularly “common antibiotics”.

“So very rarely do we find a bacteria or culture which is sensitive to a common antibiotic. Nowadays we get patient which are sensitive to…colistin, tigecycline.” (O-05).

Perceived causes of AMR development were misuse of antibiotics, namely overprescription, multiple treatments, patient non-adherence to prescribed regimens, extended duration of antibiotic therapy not based on evidence, and “easily stepping up, but afraid of stepping down” (S-12), as described below.

“Antibiotic prescription, somewhat I think, needs an improvement in our practices because though we have some established dictums in our mind that these kind of antibiotics [work], we don’t step down easily. We step up our antibiotic usage easily, we don’t step down.” (S-12).

An important contributing factor to the development of AMR, as identified by many interviewees, was that patients had already been treated with antibiotics somewhere else before coming to the hospital. Some participants placed the main fault for AMR development at primary and secondary healthcare facilities and described their practices as the random use of antibiotics by untrained personnel, indicating frequent treatment of the common cold with antibiotics. Similarly, fully private clinics were seen as contributing more to AMR than their hospitals due to a lack of supervision and specialized doctors.

“You can buy antibiotics over-the-counter in India. It is not a big thing. It’s very much irrational use of antibiotics in the peripheral zones. Patients first take the treatment in the periphery. They are getting antibiotics from the quacks and everything. They are just prescribing everything, steroids, antibiotics, antacids and everything. And then when they do not improve, they come to us. So, they have already taken the antibiotics, and we don’t know what they have taken. So, we have to start on our judgment.” (S-13).

While discussing the AMR problem, many participants compared their setting in India to high-income countries and mentioned some of the contributing factors to the high AMR burden in India, like unrestricted availability of OTC antibiotics, high production and use of antibiotics, and AMR being more of a social problem, rather than being solely caused by doctors’ treatment.

“But in countries who are developing or not developed, a country like us, India, the most important thing we should start is health promotion, first thing. Healthy practices. Until and unless we will teach the people, it is not possible to achieve this goal. We are doing many things from our part. We are doing each and everything. We are taking cultures, we are prescribing antibiotics, targeted antibiotics, broad spectrum. What not we are doing. But again, we have treated the patient, he has survived, we have made efforts, and patient has survived and now again what he will do? He will go in the same scenario. So, what will happen? He is not knowing the exact value of cleanliness and healthy practices. This is the most important thing which should be done worldwide. It is basically a social problem because a trained person will not do deliberately all these things…So on mass scale, it is very necessary to promote such practices in the remotest area of the world.” (S-08).

Surgeons reflected on AMR’s influence on patients, by saying that although most patients recover, it profoundly affects every aspect of patients’ lives. The patient’s physical state is affected by the deterioration of the condition, complications, prolonged hospitalization and treatment. In addition, it was also mentioned how AMR negatively affects patient’s mental health and social and family life. Furthermore, the big economic impact of AMR on patients was repeatedly highlighted, as “higher” antibiotics, like colistin and tigecycline, are very costly.

In terms of how AMR influences doctors, interviewees mentioned that the occurrence of AMR does not reflect well on them as healthcare professionals. Surgeons expressed a sense of responsibility for their patients, and feelings of stress, frustration, “tied hands” (O-05), hopelessness and failure when dealing with AMR, which was described as “nightmare for patients and us”(O-04).

“What can we do? Our hands are tied. We don’t give antibiotics for fun! I mean, we have to give some antibiotics…in case of an open fracture, to limit the spread of infection. If it is not infected, to limit the infection…We have to give some antibiotics. But what antibiotics? Our hands are tied. We are just part of the community.” (O-05).

“We have investigated, culture is resistant to all the antibiotics, then what should we do? We cannot sit. We cannot sit and wait and see patient to die. So, there is, because there is no guidelines at present, what should be done to such cases? Most of the time we start some higher broad-spectrum antibiotics.” (S-08).

Although most surgeons were uninformed about the extent of the AMR burden in their departments and hospitals, they often discussed the treatment of resistant cases in departmental meetings. They emphasized how resistant infections were very challenging to treat and required change of case management plan, sending repeated cultures, prescription of “higher” broad-spectrum antibiotics (e.g. carbapenems, colistin, tigecycline etc.), non-pharmacological treatment (e.g., repeated cleaning and dressings of the wound, drainage of the abscess, debridement of the bone and joint, etc.) and in some cases referral of the patient. Some participants revealed that when “higher” antibiotics were not working, they experienced “basic” older antibiotics to be effective, like sulfamethoxazole-trimethoprim.

In discussing ways to reduce AMR, a “collective effort from everybody” (S-13) approach was suggested. The need to improve IPC measures was highlighted by improving both healthcare staff’s IPC practices and the overall cleanliness of hospitals. Additionally, optimizing antibiotic use was emphasized, with special attention to issues like unindicated prescribing, inappropriate dosing and treatment duration, standardization of perioperative surgical prophylaxis, and promoting strategies such as de-escalation, regular prescription audits and feedback, and more frequent use of early-generation and topical antibiotics. A major recommendation was to develop department-specific antibiotic prescribing guidelines that are regularly updated to reflect local resistance patterns and include more cost-effective antibiotic options. Most participants stressed the importance of regular departmental meetings on AMR, training programs for doctors and medical staff, and collaboration with microbiology and pharmacology departments. Junior residents often pointed out that senior consultants should be regularly updated on new protocols and guidelines to endorse their implementation. Stricter regulation of antibiotic use was also seen as necessary, at the hospital level through stricter protocols, and at national level through tighter controls on antibiotic prescriptions in primary care and limits on OTC sales. Finally, the need for “shift in the patients’ mindset and behavior” (O-05) was emphasized, through public education and health campaigns focused on hygiene, antibiotics and AMR.